Genetic Diversity and Interspecific Relationships in Banksia L.F., (Proteaceae)
Genetic Diversity and Interspecific Relationships in Banksia L.f., (Proteaceae).
Tina Louise Maguire B. Ag. Sci. (Hons)
Submitted in fulfrlment of the requirements for the degree of
Doctor of Philosophy
Deparrnent of Horticulnue, Viticulture and Oenology
University of Adelaide
August 1996 Baril Table of Contents Abstract i Declaration and authority of access to copying iv Acknowldgments v List of Tables vi List of Figures ix Chapter One: General inroduction 1 The genus Banksía 1 Thesis aims 3 Chapter Two: Literature review 5 Introduction 5 Taxonomy 5 Conservation 8 Plant population management 10 Breeding Systems 11 B anl Alternative methods of hybridisation 30 Hybrid Verification 31 Pollen Collection and Uses 32 Factors Controlling Pollen Availability 32 Genetic controls 32 Extemal controls JJ Pollen Storage 33 Relative humidity (RH.) 34 Temperature 34 Gas atmosphere and oxygen pressure 34 Causes of decreased viability during storage 35 Viability tests 35 In vitro assays 36 Non germination assays 36 Recording data 37 Comparison of viability tests 37 Molecular Techniques to Study Relationships in Plants 38 The poly'rnerase chain reaction (PCR) technique 38 Random amplified polymorphic DNA (RAPD) 39 DNA sequencing 4T Molecular Approaches to Plant Systematics 43 Chloroplast DNA M PCR analysis of chloroplast DNA 45 Analysis of DNA sequence data 46 Conclusions 46 Chapter Three: Viability testing of Banl Effect of position on the inflorescence and time during the flowering season 52 Pollen storage 52 Discussion 60 Chapter Four: Interspecific and intergeneric pollination with Banksia coccinea R.Bt (Proteaceae) 62 Abstract 62 lntroduction 63 Materials and methods 66 Plantmaterial 66 Pollinations for assessment of pollen tube growth 66 Resuits 67 Effect on pollen tube growth of site, month and year 67 Reciprocal crosses 67 Interspecific and intergeneric pollination with,B. cocc[nea as male parent 68 Pollen tube abnormaliúes 69 Discussion 81 Chapter Fíve Banksi¿ sect. Coccinea (4.S. George) T.Maguire et aI., (Proteaceae). A new section 86 Abstract 86 Introduction 86 Materials and methods 88 B anl Chapter Six: Seed set following interspecific pollination with B. coccinea R.Br. 95 Abstract 95 Introduction 95 Materials and methods 96 Plant material 96 Interspecific seed set 97 Germination trials 97 Hybrid verification 98 Table of Contents Results 98 Interspecif,rc seed set 98 Seed germination of B. coccinea, B. ericiþlia and B. coccinea x B. ericiþIia 99 . Seedling survival and vigour 99 Hybrid verification and early seedling cha¡acters 99 Discussion rt2 Chapter Seven: DNA isolation methods for Banksia and other members of the hoteaceae lr4 Abstract tL4 Introduction 114 Materials and methods 115 Results 116 Protocol 1: DNA extraction from mature leaves 116 Protocol 2: DNA extraction from seedling leaves and other material high in phenolics and polysaccha¡ides 118 hotocol 3: Isolation of toul genomic DNA from seed material (slightly modified "miniprep" method of Weining and Langridge, 1991) lzl Gel electrophoresis I22 Discussion L22 Chapter Eight: Genetic diversity of Banksia and Dryandra (Proteaceae) using RAPD markers. 123 Abstract L23 Introduction I23 Materials and methods 125 Plant material 125 DNA extraction I25 DNA amplification and documentation 126 Data analysis lZ7 Results l2g Discussion I32 Chapter Nine: RAPD variation within and between populations of Bøn*sia cuneata A.S. George (Proteaceae), a rare and endangered species. 134 Abstract 134 Introduction I34 Table of Contents Mate¡ials and methods 136 Population sampling 136 DNA isolation 136 DNA amplification and documentation r37 . Statistical analysis 138 Results r39 The RAPD profile t39 Estimate of genetic diversity r39 Discussion 148 Chapter Ten: Use of RAPD ma¡kers to analyse phylogenetic relationships in Banksta (Proteaceae) 153 Abstract 153 Inroduction 153 Materials and methods 155 Plant material 155 DNA extraction 155 DNA amplification and documentation 156 Data analysis I57 Genetic distance analysis I57 Phylogenetic analysis using parsimony (PAUP) I57 Results 158 Discussion 166 Chapter Eleven: Apptication of non-coding chloroplast DNA sequences to Banksia (Proteaceae) phylogeny 168 Abstract 168 Inroduction 168 Materials a¡d methods 171 Plantmaterial I7I DNA extraction l7I DNA amplification I72 DNA sequencing 172 Data analysis I73 Results 173 Discussion 176 Chapter Twelve: General discussion 179 Banlcsia breeding 179 Table of Contents Genetic diversity 181 Species relationships 184 References cited 187 Appendix 219 Abstract Abstract Banksias are amongst the best known Australian wild flowers. They are used in ornamental horticulture and last well as fresh cut flowers, or indefinitely as dried arangements. Breeding and selection of new cultiva¡s for the cut flower industry is currently underway. This thesis aims to increase knowledge essential for conservation biology and for focused and efficient breeding of banksias. Pollen storage and viability testing are important adjuncts to a plant breeding prog¡am. 0C Banksia menziesii pollen was stored at 20 0C, 40C, -20 0C, -80 0C and -196 and assessed using a semi solid medium of IVo agar, L57o sucrose, 0.0l%o boric acid, 0-037o calcium nitrate, 0.02V0 magnesium sulphate, 0.017o potassium nitrate, and an incubation temperature o125 0C. Germination remained constant at afoundTÙVo in all treatments except room remperature 20 0C, which after six months had only 257o getmination. Pollen viability was assessed using fluorescein diacetate (FDA), but the results did not reflect the loss of germinability at 20 0C. There was no effect of floret position on the inflorescence on germination; but pollen viability va¡ied over the flowering period with maximum germination mid season. Interspecific hybridisation is assessed as a potential breeding tool, and for the assessment of species relationships within the genus. Pollen tube growth was investigated using controlled hand pollination of the commercialty significant species Banl species of Banlcsia, and the related genus, Dryandra. Currently, the relationship between B. coccinea and the other species groups within Banksia is unclear. It has been found previously that success of pollen tube growth in the pistil following interspecifrc pollination was largely related to trxonomic distance between the species (Sedgley et al. lg94). Thus, interspecific hybridisation is a suitable technique to determine the compatibility relationships of the problematic species B. coccíneø. Some species supported no germination of B. coccinea pollen. Others produced pollen tube 1 Absnact abnormalities including thickened walls, bulbous swellings, non-directional growth, burst tubes and. branched tubes. Control of pollen tube growth in the pistil was imposed in the pollen presenter, a specialised region of the style for pollen presentation to foraging fauna, and in the upper style. There was no significant reciprocal effect on pollination success in the lower style. The results of pollen tube compatibility in the lower style indicated that B. coccinea had a closer affinity to the section OncosryIis, than to secúon Bant Banksía to a new section Coccínea, the sister section to Oncosrylls. Intergeneric crosses of B. coccínea with Dryandra species resulted in some compatibility, with one cross having low numbers of polien tubes in the pollen presenter and upper style region. These results indicate a close relationship between Banksia and Dryandr¿, which a¡e sister genera in the ttbe Banksiae,farnily Proteaceae. Species relationships within Banksia were also assessed using molecula¡ techniques. Random amplified polymorphic DNA (RAPD) markers were assessed for their usefulness at various taxonomic levels within the genus. It was found that RAPDs a¡e informative at the close species level, but not at more distant levels, such as between distantly related series, sections, and subgenera. In addition, species relationships at higher levels were investigated using direct polymerase chain reaction (PCR) sequencing of chloroplast DNA (cpDNA) spacer regions between the trnL and trnF exons. These regions are thought to be universal for plant species and informative at the intra and interspecific level of plants. Using the region be¡ween trnL and nnF, relationships within Banl region was conservative, with little variation between species. Section Banl group, section Oncosrylis formed another group, and B. coccinea along with two Dryand,ra species was placed between the two sections. Resolution at this node however, was not complete. Subgenus Isostylis formed two groups away from the two sections in subgenus Banksia,withB. illiciþIia andD. formosa together, while B. cuneatawas 11 Atlstrâct more distantly related. Based on DNA sequence and RAPD data, it appears that Banksia and, Dryandramay be a¡tificial genera, and ttrat in the presence of each other, they cannot be separated using RAPD or trnLDNA sequence data. Genetic variability within species of. Banksia was investigated using RAPDs. Levels of genetic diversity were generally high, ranging from 0.59 - 0.90. This agrees with previous work using isozymes, pollen tube and fruit set data, showing that Banksia species are predominantly outcrossing. In particular, a detailed study was conducted on a geographically restricted, rare and endangered species, B. cuneara. Using RAPDs on all known populations, it was found that levels of genetic diversity were high, ranging from 0.65 - 0.74, andthat there was no significant genetic differentiation between populations. In conclusion, this study contributes to knowledge essential for further improvement and conservation of Banksiø species, and raises questions regarding the currently accepted taxonomic relationships within Bø nl ru Declaration and authority of access to copying This work conrains no material which has been accepted for the award of any other degree or diploma in any university or other tertiary institution and, to the best of my knowledge and betief, contains no material previously published or written by another person, except where due reference has been made in the text. I give consent to this copy of my thesis, when deposited in the university library, being available for loan and photocopylng. SIGNED DATE: l3c) i Ê/c(óì lv Acknowledgments This work was conducted in the Department of Horticulture, Viticulture and Oenology, University of Adelaide, with the support from an Australian Postgraduate Research Award, and partial support from the International P¡otea Association. I would like to sincerely thank my supervisors Prof. Margaret Sedgley, Dr. John Conran and Dr. Graham Collins for guidance, encouragemsnt, support, time and dedication. Thanks also to SA Water and Phil Howlett for access to Banksía collections; Lynne Giles, Rita Middelberg and Trevor Hancock for statistical assistance; Dr. Rod Peakall for assistance with the RAPDistance package and Dr. Laurent Excoff,rer for providing WINAMOVA; Alex George and Kevin Thiele for discussion; Department of Conservation and Land Management, WA, for a permit to collect B. cuneata seed, Leon Silvester for assistance in seed collection, Dr. David Coates for discussion; Dr. B. Hyland, CSIRO, Atherton, Qld, for Musgravea specimens; Dr. Neil Shirley and Jing Li for assitance in DNA sequencing; Michelle Wirthensohn and other fellow postgraduate students in the department for their friendship and help; my family for their support and encouragement, and Stephen Jiannis for his support and keeping me sane. v List of Tables List of Tables Table 2.1 Classification within Banlcsia (George 1987). 7 Tabke 2.2 Recorded hybrid banksias. 23 Table 3.1 Soil profile of Happy Valley and Nangkita field sites. 53 Table 3.2 Optimisation of in vitropollen germination medium forB. menziesíi. 54 Table 3.3 Effect of position on the inflorescence and time during the flowering season on viability of pollen of B. menziesü. 55 Table 3.4 Effect of storage temperature and time on viability of pollen of B. menziesü 56 Table 3.5 Correlation coefficient beween FDA staining results andin vino germination percentage for pollen storage over six months. 57 Table 4.1 Site effect on mean pollen grain and tube numbers in interspecifrc crosses of Banlcsia with B. coccinea as the male parent. 70 TabIe 4.2 Effect of time of pollination during the flowering season on mean pollen grain and tube numbers in intraspecific crosses of Banlcsia coccinea and B. menziesü 77 Tabte 4.3 Mean pollen grain and tube numbers in interspecif,rc crosses of Banl Table 4.4 Mean pollen grain and tube numbers in intraspecific, interspecific, and intergeneric crosses pollinated with B. coccineapollen. 73 Table 4.5 Predicted probability of pollen tube growth in intraspecifrc, interspecifrc and intergeneric crosses pollinated withB.coccineapollen- 74 Table 4.6 Mean pollen grain and tube numbers in intraspecific, interspecific and intergeneric crosses pollinated with B. coccinea pollen averaged over taxonomic groups. 75 vl List of Tables Table 4.7 Predicted probability of pollen tube growth in intraspecific, interspecific and intergeneric crosses pollinated with B. coccinea pollen averaged over taxonomic groups. 76 Table 4.8 Pollen tube abnormalities observed in interspecific crosses with B. coccinea pollen. 77 Table 5.1 Pollen tube penetration of interspecif,rc and intraspecific pollen in Banksia pistils pollinated with B. coccíneapollen. 91 Table 6.1 Number of fertile cones and mean number of follicles/cone following interspecific pollination with B. coccineapollen (SE = standad error). 101 Tabte 6.2 Mean follicle and seed set data for parental species and putative interspecifrc hybrid B. micrantltaxB. coccinea . 102 Tabte 6.3 Mean folticle and seed set data for parental species and putative interspecific hybrid B. ericiþIiaxB. coccinea. 103 Table 6.4 Seed germination percentage and germination time for parental and putative hybrid seed. 104 Table 6.5 Seed.ling morphology data for 3 month old seedlings of B. coccinea, B ericiþlia, and presum ed, B. ericiþIia x B. coccinea hybrid. 105 Table 6.6 Morphological cha¡acter evaluation. 106 Table 8.1 Band data and estimates of diversity for 33 species of Banksia and three of Dryandrausing RAPDs (standard deviation). 129 Table 8.2 Genetic diversity of three species of Banksia calculated from RAPD data using six different merhods compared with published results from allozyme analysis. 130 Table 9.1 Summary of data obtained by RAPD analysis for 5 primers with 125 individuals of B. cuneata. l4I Table 9.2 Summary of band frequencies for each population of B. cuneata. 142 vll List of Tables Table 9.3 Genetic diversity of each population of B. cuneata calculated using the simitarity (F) metric of Nei and Li (1979), where distance (D) = 1 - F. 143 Table 9.4 Analysis of molecula¡ variance (AMOVA) for 125 individuals of B. cuneata using 169 RAPD bands. IM Table 10.1 Primer sequence and band data for each of the 15 selected primers over the complete data set of 37 taxa producing 791 bands. 160 Table 10.2 Band data for 33 species of Banksia, three of Dryandra and Musgravea heterophyllausing RAPDs. 161 vtll List of Fisures List of Figures Figure 3.1 Temperature and rainfall data recorded at the nearest weather station to the Happy Valley experimental site for 1993, 1994 and 7995. 58 Figure 3.2 Temperature and rainfall data recorded at the nearest weather station to the Nangkita experimental site for 1993, 1994 and L995. 59 Plate 4.1 Interspecifîc hybridisation technique for Banl Plate 4.2 Scanning and fluorescence micrographs of Banlcsia. 79 Ptate 4.3 Pollen tube abnormalities observed inBønksia coccinea interspecific crosses using fluorescence microscopy. 80 Figure 6.1 Schematic drawing of parental and putative interspecific hybrid seed 107 Figure 6.2 Putative interspecific hybrid seedling between B. ericiþlía and B . coccinea at 3 months. 109 Figure 6.3 Schematic drawing of parental and putative interspecific hybrid cotyledons. 109 Figure 6.4 Dendogram showing relationships between B. coccinea,B. ericiþlía and putâtivo interspecific hybrid. 110 Figure 6.5 Multidimensional scaling of data matrix into two dimensions (stress value 0.05). 111 Figure 7.1 Agarose gel (L.6Vo) showing DNA yield of Banksia cuneata individuals following DNA extraction and RNAase digestion. 120 Figure 8.1 Agarose gel showing RAPD markers produced using primer OPB-7 for 10 individuals of B. ashbyi. l3I Figure9.1 Geographic range of Banksia cuneata in south-west Australia, showing locations of the ten populations from which seed was collected. 145 D( List of Fieures Figure 9.2 Agarose gel showing RAPD markers produced using primer OPA - 9 for individuals of B. cuneata. r46 Figure 9.3 UPGMA cluster analysis of B. cunearø populations based on genetic distance (PhiST) 4rnong populations calculated by AMOVA analysis. I47 Figure 10.1 Agarose gel showing RAPD ma¡kers produced using primer OPB - 11 for bulked DNA samples of 34 species. 162 Figure 10.2 Dendogram generated by cluster (UPGMA) analysis of genetic distance values generated from 791 RAPD bands using 15 primers. 163 Figure 10.3 Majority rule consensus tree obtained f¡om PAUP, tree length = 4079; consistency index =0.929. L& Figure 10.4 Strict consensus tree obtained from PAUP, tree length = 4079; consistency index = 0.929. 165 Figure 11.1 Complete nucleotide sequences of the spacer between the trnL (UAA) and rrnF (GAA) gene inBanksia,DryandraandMusgravea(lengthof alignment4l3bp.lT4 Figure 11.2 Phylogenetic relationships within Bartksia and related genera. 175 x General introduction Chapter One General introduction The genus Banksia The genus Banlcsia L.f., which belongs to the family Proteaceae, is named after Joseph Banks who collected the first specimens in 1770. Banksias are amongst the best known Australian wildflowers, with approximately 75 species (George 1987), all of which are native to Australia. The greatest concentration of species is in south west Western Australia where 60 species a¡e found. Fourteen occur in south east and eastern Australia from Eyre Peninsula to Cape York Peninsula, and two of these also occur in Tasmania. One tropical species, B. dentata, extends from the Kimberley to Cape York Peninsula, and is also found in Papua New Guinea, kian Jaya and the Aru Islands. Apa.t from B. dentata no species is common to both east and west Australia, but taxonomically there arc strong relationships between the two groups. The geographical range of. Banltsia species varies greatly. Some are widespread and common, such as B. príorntes, whgreas others are very restricæd such as B. cunearø (George 1987). Species relationships within genus Banksia are not fully resolved, and some such as B. coccinea are problematic and have no known relatives. B. coccineø is unusual as it has unique leaves, inflorescences, pollen, infructescences, follicles and seeds (George 1987). A recent cladistic analysis of Banksia using morphological cha¡acters (Thiele t993) confrrmed the classification of George (1981, 1988) for some species, but not for others. B. coccinea was placed incerte sedis. Other additional independent data are required to resolve these relationships. Higher order relationships between subgenus Banksia, subgenus IsosryIís and genus Dryandra are also unresolved. Banksía and Dryandra ue I General inroduction sister taxa, in the tribe Banksiae, family Proteaceae, and are thought to show parallel development. Variability within genus Banksia and within species is high, with a wide range of forms, growth habits, morphology, and environmental tolerances. The genus represonts a large source of genetic variability, only a portion of which is being utilised in ornamental horticulture. Until recently flowers were ha¡vested from the wild, but now there is commercial planting of banksias for the cut flower industry. Breeding and selection is underway (Sedgley et al. L99t,7994), for improved cultiva¡s for the cut flower industry. Selection of improved va¡ieties from variable populations is important for future breeding efforts, so conservation of natural variation in Banksia is essential. This ensures suffrcient resources to produce novel hybrids, and improved growth rates, plant form, local adaptation, disease resistance, yield, bloom colours and other important horticultural characters. Effective conservation and management of natural stands depends upon sound knowledge of the biology and population genetics of species. Issues affecting Banksia population viability include habitat fragmentation, land clearance, altered fire and recruitrnent frequencies, disturbance, and introduction of disease and non indigenous plants and animals. Further research into levels and patterns of genetic variation will result in informed decisions, to ensure the viability of Banksia populations. 2 General innoduction Thesis aims This project aims to increase knowledge essential to Banksia breeding and conservation. Interspecific hybridisation is assessed as a potential tool for the production of novel hybrids. Pollen storage is investigated to increase flexibility of the interspecifrc breeding program. Species relationships within the genus a¡e investigated; this information is essential for focused breeding efforts. Genetic diversity within species of Banksia ís investigated for conservation of genetic resources, and patterns of genetic diversity within and between populations of a rare and endangered species B. cuneata are investigated. Details of each project aim are given below. P oIIe n vîabilíry and storage Pollen srorage and viability testing is an important component of plant breeding progrÍrms, when the species to be hybridised do not flower at the same time or are geographically isolated. Flexibility of the breeding program will be $eatly increased if pollen can be storcd successfully for extended periods. There is no published information on pollen collection and storage, or on invitro conditions for successful pollen germination and tube growth of Banksía species. The aim of this study is to develop ín vitro viability testing methods for Banlcsiapollen and to develop practical methods for long term storage. Species relationships Species relationships are important for classification and have implications for the evolution of a group. If species relationships are resolved, then breeding and introduction of foreign germplasm for crop improvement can be more focused and efficient. This project investigates species relationships in two ways, through interspecific hybridisation and via molecular techniques. Potential crosses identified using interspecific hybridisation will be investigated furttrer for seed set and new cultivar development. Various molecular methods ^5 General introduction can be used to address different questions. Random amplified polymorphic DNA (RAPD) markers are assessed as a potential tool for phylogeny within genus Banksia. Another technique, direct polymerase chain reaction (PCR) sequencing of chloroplast DNA, is used to investigate higher order species relationships within genus Banlcsia, between subgenus Banksía, and subgenus lsosfylis, and between genus Banskia and genus Dryandra. Currently, relationships beween these genera and also within genus Banl Genetic diversiry in rørural populations o/Banksia The breeding system and levels of genetic diversity in Banksia species are important in population genetics, conservation and management of endangered or threatened species, and ulúmate species survival. To date, there is little information on levels of genetic diversity and breeding systems of Banksia species, including many rare or threatened species. This project investigates levels of genetic diversity for a number of species of Banlcsia using random amplified polymorphic DNA (RAPD) markers to estimate genetic diversity in natural populations of Ban|csía. Patterns of genetic diversiry within and between populations of B. cuneata, a rare and end.angered species Genetic diversity within a¡d between all known populations of B. cuneata is investigated. B. cuneata is declared a rare species under the V/A V/ild1ife Conservation Act, and is under threat of extinction due to small population sizes and fragmented distribution over private and public land. Genetic studies including all known remaining populations have not been conducted. Levels and patterns of genetic diversity are investigated using RAPDs, and this information is important for future conservation and management strategies forB. cuneata. 4 Chapter Two Literature review Introduction The biology of Banksia breeding systems is important in the context of interspecifrc pollination, pollen storage, species development and isolation, species relationships, classification, and conservation of natural resources. In addition, related topics such as genetic markers in plant breeding, population genetics and phylogeny make an important contribution to B anksia improvement. Taxonomy The first species of Banlcsia to be described (8. serrata, B. integriþlia, B. dentata andB. ericíþIia) were collected on the east coast of Australia at Botany Bay and Endeavour Bay by J. Banks and D. Solander in 1770. Ca¡l Linnaeus named Banksia based on these four species in 1781. Subsequently, there have been other species described by Brown (1810, 1830), Meissner (1855), Gardner (1928, 1964) and George (1981). Since the monograph of George (19S1) only three further taxa have been described. The cu:rently accepted classifrcation of George (1987) is shown in Table 2.1. Early treatments of Banksia were either catalogues of species (Brown), or artificial classifrcations (Meissner). George attempted a more natural classification, using a variety of leaf, flower, follicle and seed characters to base infrageneric taxa. While George describes "possible lines of evolution" within and between taxa, his classification is difficult to interpret phylogenetically because of a number of taxa with uncertain relationships. 5 Literahuereview Doust (1983) attempted a cladistic analysis of Banksia using 37 chancters for 86 taxa, but the trees were poorly resolved. This analysis provided a poor match with the existing taxonomy and failed to suggest a robust phylogenetic scheme. It was not used for a formal classification. More recently Thiele (1993) proposed a classification based on cladistic analysis of morphologlcal cha¡acters. Although there is some agrcement with the currently accepted classification of George (1987), there are still some uncertainties in the positions of some species and support for some nodes is tenuous. length Venkata Rao Relationships wittrin the family Proteaceae have been discussed at by 1 (1971), Smith-White (1959), Ramsey (1963) and Johnson and Briggs (L975,1981). The views of Johnson and Briggs reprcsent the most recent and accepted classification for the family. The Proteaceae is a large and diverse family of trees and sh¡ubs comprising 75 genera and 1500 species (Johnson and Briggs 1983), found in Australia, South Africa and South America, and is presumably Gondwanan in origin. Few genera are common to South America and Australasia with most genera and species endemic to one continent. Fossil records indicate a very ancient family dating back to the mid to late Cretaceous and Tertiary period (Dettman 1973, Martin 1978). Relationships amongst extant Proteaceae are tentative with Banksía placed in the subfamily Grevilleoideae. V/ithin this subfamily lrrbes Bo"rksrect¿ there are two suiìËåbes, Éàrikisielc and Musgraveinae, the latter restricted to the bonksiina¿ rainforests of north oastern Australia. Within subtribe Ba¡ldsieae there are two genera Banksia and Dryandra. All authors agrce that Banksía and Dryandra are each well defined, natural and monophyletic gloups, although there are some parallel developments within the two genera. Banlcsia comprises two subgenera,IsosryIis and Banlcsia. The former group is small and le¿ the species are simila¡ to Dryandra. This Gãä George (1981) to propose that /sosry/is may be closer to Dryandro than to subgenus Banksía. Nevertheless, he placed it as a ó Liæran¡rereview subgenus of Banlcsía. with the comment that a separate genus may be appropriate. This Çouc interpretation was based on dlrree morphological characters. The three species in rn volr* c cc.ì subgenus Isosrylis lack the prominent invoftd bracts typical of Dryandra, but have flower subtending bracts typical of Banksiø, their follicles are thick, woody and tomentose as in Banlcsia and they have the ovoid inflorescence axis typical of Banksía. There is still speculation however, on the relationships of subgenus Banl Isostylis, and genus Dryandra.It is possible that other data such as DNA sequence, or .o¡196Cq\ ìve otherDNA eemþa*ise"methods may resolve relationships at this level. Table 2.1 Classification within Banlæía (George 1987) SUBGENUS BANKSIA Section Banksia Series Salicinae: B. canei, B. conferta, B. dentata, Series Cyrtostylis: B. ashbyi, B. attenuata, B B. integriþlia, B. marginata, B. oblongfolia, B. audax, B. bentlnmiana, B. elderiana, B. elegøns, B paludosa, B. plagiocarpa, B. robw, B. saxicola epica, B. løevigata, B. lindleyana, B. Iullfttzü, B tnedia, B. pilostylis, B. praemor sa. Series Grandes: B. grandis, B. solandri. Series Prosratae: B. blechnifolia, B chamaephyton, B. gardneri, B. goodii, B petiolaris, B. repens Series Quercinae: .8. oreophila, B. querciþlia Series Tetragonae: B. aculeata, B. caleyi, B Iemarni¿na Series Bauerinae: B. bau¿ri Series Coccineae: B. coccinea Series Banksia: B. aemula, B. baxteri, B. Section Oncostylis candolleana, B. menziesü, B. ornata, B. scepffum, Series Spicigerøe: B. brownü, B. ericifolíø, B B. serrata, B. speciosa. Iittoralis, B. occidentalis, B. seminuda, B spinulosa, B. tricuspis, B. verticillata Series Crocinae: 8. burdettü, B. hookeriana, B. Series Dryandroideae: B. dryondroidcs. prionotes, B. victoriae. Series Cyrtostylis: B. ashbyi, B. attenuata, B. Series Abietinae: .8. grossa, B. incana, B. lanata, audax, B. bentha¡niana, B . elderiana, B. ele gans, B. B. laricina, B. leptophylla, B. meisneri, B. epica, B. laevigata, B. lindleyana, B. lulfitzü, B. micrantha, B. nutans, B. pulchella, B. scabr ella, B . media, B. pilostylis, B. praemorsa. sphaerocarpa, B. telmatiaea, B. violacea. SUBGENUS ISOSTYLIS B. cuneata. B. ilicifolia. B. olipantln. 7 Liæranuereview Conservation The reasons put forward for the preservation of species are va¡ied. They range from the belief that all plants and animals have the right to exist, to the view that future generations have the right to expect current resources. Strict preservationism is not the same as conservation. Conservation may allow preservation of species and ecosystems, but also allow their use in a way that is not darnaging or wasteful. P¡eservationism is concerned with preventing destruction of species, ecosystems and wilderness, when they may be regarded as resources by others. To adopt a preservationist view would result in world- wide sta¡vation, and to have uncontrolled consumption would lead to the same through exhaustion of resources. A challenge to the conservationist is to seek middle ground when promoting conservation and controlled exploitation. Extinction rates for flora and fauna are increasing, through environmental change caused by habitat destn¡ction, introduced weeds and animals, and pollution. Threats to plants include agriculnue, urban development, mining, quarrying, chemicals, overgrazing by animals, cultivation, fire, disease and disappearance of pollinators. Extinction of a species involves a reduction in numbers until it ceases to exist. Extinction is dynamic and occurs when environmental conditions exceed the adaptive capacity of the individuals. Two classes of events lead to extinction, deterministic and stochastic events. Deterministic events are those which involve unalterable change such as deforestation or climate shift. Stochastic events are random and are a result of normal changes or environmental disturbance. Stochastic events reduce population size and density and make them more susceptible to other events. In nature, many extinctions are brought about by deterministic events which reduce numbers and then stochastic events a¡e influential. High environmental variability is the greatest ba¡rier to persistenco, rather than demographic and genetic uncertainty. E Literatu¡e review Stochastic forces fall into three categories. Demographic stochasticity, which is the chance va¡iation of low births and high deaths. The second type is genetic stochasticity, which includes inbreeding depression, chance effects of lethal genes, and the loss of genetic diversity. and heterosis. The third type is envi¡onmental, and it includes environmental shocks felt by all members of a population. In a dispersed population there is likely to be more habitat va¡iation that there are buffers against change. However, no environment is static and consequently all populations are liable to environmental stochasticity. Deterministic events are readily observed such as storms, fires and feral animals, but stochastic events may go unnoticed until it is too late. This is why monitoring of threatened populations is important for conservation of species. In each case of potential extinction specif,rc strategies are required to protect the species, based on life history, distribution, habitat, fluctuations in predator relationships and inherent genetic variation. However, in general it can be said that as population size decreases the threat of extinction increases. In the very long term, survival and persistence depend on tho adaptability of a population to change its genetic composition in response to long term changes in the environment. This depends on genetic resources available, hence the need for large effective population sizes. Effective population size is not just the total number of plants. Effective size may be reduced by the presence of non breeding individuals, skewed sex ratios, inbreeding, random variation in progeny, loss of genetic diversity through a bottleneck period of low population size and patterns of pollination. Some populations may be far too small to survive even if they appear to be healthy and viable. Such species may be on the way to extinction unless there is rapid intervention. Natural populations of Banlcsia are subject to a number of pressures such as land clearing for agriculture, mines, roads, exploitation for cut flowers, burning a¡d disease. In Western Australia (WA) clearing threatens the survival of B. cuneatø and B. hookeríana. 9 Liæratuereview Populations of species that are killed by fire and regenerate from seed, may not survive two fires in quick succession since they would not have enough time to mature, flower and produce more seed. Rare species that would suffer in this respect are B. aculeata, B. burdettä, B. conferta, B. cuneata, B. dryandroides, B. lanata, B.laricina, B. meisneri, B. pilostylís, B. praemorsa, B. querciþIia, B. scabrella, B. telmatiaea, B. vericillata and B. victoriae (George 1987). Many species are exploited commercially including B. coccinea, B. hookeríana, B.prionotes, B.menzíesü, B. baxteri and B. speciosø. V/ith continued decline in inflorescence number there may be a reduction in the seed bank and subsequent loss of genetic variability. Under the WA V/ildlife Conservation Act 1950-1979 there are seven Banlæias species that are declared rare flora. They are B. brownií, B. cltamaephyton, B. cuneata, B. goodii, B. meisneri, B. sphaerocarpa andB. tricuspís. It is essential that all species of. Banksia be conserved in the wild. Banksias are important components of the vegetation in which they occur and are a food source and host for many species of native fauna (George 1987). They are also a propagation source and tourist attraction. Native plants are Australia's greatest horticultural asset and there is a need to safeguard germplasm for the sake of future production and education. P lant p opulatio n mnnog e me nt Many aspects of plant population biology must be considered during the development of conservation snategies. These include: breeding systems, life cycle patterns, dispersal modes and their application to management. Species management involves management of particular populations at specifrc sites, but species conservation cannot be considered to be independent of habitats and ecosystems. The use of land adjacent to the site is relevant to the conservation of particular populations. Social issues are also important, as the particular site or species may be of interest to several groups whose goals may be incompatible with conservation. There a¡e also likely to be budget and staff constraints. Sometimes effective management simply consists of an initial survey, followed by 10 Literah¡re ¡eview protection of the site from unwanted human interference, with periodic visits to check the state of the habitat and number of plants. If intervention is necessary, the management choices multiply, and selection of the appropriate method may become difficult. Population management needs to be considered from the following points of view: (1) Ecosystem management depends on maintenance of key species, whose decline could affect the whole ecosystom. (2) Particular species may be threatened because of factors outside the ecosystem, such as harvesting and the consequences of small population size. (3) Some species may require specific habitat manipulation. (4) It cannot be assumed that because a species is present now, that it will still be present in 50 or 100 years time. Strategies need to consider the expected persistence time. (5) People generally frnd it easier to understand conservation of biodiversity at the species level rather than at the ecosystem and gene levels. Species conservation is best done at the population level. (6) The ultimate test of management success is whether it promotes the persistence of biodiversity into the future. Given adequate protection and assistance in reproduction, many rare and endangered plants can recover in numbers and vitality. The role of conservation biology is to shift the balance in their favou¡. Breeding Systems V/hich gametes are actually brought togethü to form zygotes, depends on the breeding system of the plant. There are many floral, physiological and genetic mechanisms that contribute to the mating system of plants. It is essential to have an insight into the breeding system to understand the complex patterns of variation within and between nanral populations. There are three basic breeding mechanisms, outbreeding, inbreeding and apomixis. Most higher plants are hermaphrodite and have male and female parts, suggesting self fertilisation. It is interesting however, that many plants are adapted to cross fertilisation, and minimise or prevent self fertilisation. In most cases self incompatibitity is involved. Species that are self compatible may be able to flower and fruit in the absence of pollinators, and the progeny are often uniform in appearance. In 11 Liæranrrereview apomixis, reproduition is achieved without fertilisation and the sexual process is avoided. Two types of system are found, vegetative apomixis and agamospermy. Vegetative apomixis is the radial $owth by rhizomes, stolons and runners, and is characteristic of many perennial species. Plants arising from these propagules are the same genotype as the parent. Agamospermy is when seed is set but no sexual fusion has occur¡ed. Offspring are the s¿rme as ttre plant that produces them. The breeding system of many banksias is still unclear. Information has come from studies of seed set, pollen tube growth, and more recently enzyme electrophoresis @uss and Sedgley 1991, Carthew et al. 1988, Coates and Sokolowski 1992, Sampson er a/. 1994). High levels of outcrossing have been reported in Banksia, and are among the highest recorded for plants (Scott 1980, Schemske and Lande 1985, Carthew et aI. 1988, Coates and Sokolowski 1992). Banksias are predominantly outcrossing, although some studies have shown a mixed mating system (Fuss and Sedgley 1991). B anksia /ow er morpho Io gy Banksias produce inflorescences composed of numerous hermaphrodite flowers clustered in pairs around a central woody core. Each floret has a single elongated pistil with two ovules in the unilocular ovary. Anthers are attached near the tip of each of the four perianth parts by short filaments. This region of the perianth, prior to anthesis, encloses the distal portion of the style which in many species has an obliquely terminal stigmatic groove. This portion of the style is modified for the special function of pollen presentation. The anthers dehisce prior to anthesis depositing pollen onto the pollen presenter. At anthesis, the style is released from the perianth and pollen is available to foraging insects. The stigmatic groove becomes receptive after the pollen has been removed, when it is ready to receive pollen from another plant. Following fertilisation, bottr follicles and seeds may develop, taking from a few months to two years to mature. T2 Literahrereview Floralinitiation Floral initiation and development in relation to the time of flowering has not been widely studied in Banksía, although investigations have been conducted on B. baxteri and B. hookerianø (Rohl et al. 1994) and B. menzíesii and B. coccíneø (Fuss and Sedgley 1990). Floral initiation occurred in spring in B. baxteri and in early summer inB. hookeríana. Floral development was rapid in B. baxterí taking three months to reach anthesis, with five months in B. hookeríana, which flowered in winter. In both sPeciey shoots that flowered were thicker in thei¡ fust year than those ttrat did not. Floral initiation in B. menzíesii and B. coccinea occurred in late spring, but the rate of development was significantly different. Flowering times in Banl with a peak in autumn (Collins and Rebelo 1987). Species vary considerably with regard to the time taken for all flowers on an inflorescence to open, from four days in B. occidentalis to 57 days in B. spinulos¿ (Collins and Rebelo 1987). This could be due to factors such as inflorescence size, and period of nectar production required for pollinator attraction. The flowers at the base of the inflorescence open first in most, but not all, species of Banl termination of the growth of the perianth, leads to style curvature, such that the number of flowers opening on an inflorescence is increased by the triggering action due to contact of foraging animals such as birds. Va¡iable amounts of pollen are presented on freshly opened flowers, and counts of 1 x 103 m 3 x 103 have been reported for five species of Banlcsía (Collins and Rebelo 1987). V/ithin communities, opportunities for hybridisation among species may be minimised by staggered times of flowering, different pollen vectors and possibly by interspecifrc incompatibility or incongruity. Birds and insects generally forage in such a way as to facilitate both outcrossing and selfing. 13 Liærahrereview Protandry Protandry occurs when the male reproductive organs are mature before the female, and many genera of the Proteaceae are protandrous. It has been reportedinBanksia (Carolin 1961, Venkata Rao lgTL,Johnson and Briggs 1975) but there have been few estimates of the exact timing. In B. menziesíí maximum pollen germination and pollen tube growth occurred at 3 days posr anthesis (Fuss and Sedgley 1991). The development of maximum stigma receptivity to pollen, as indicated by secretion of esterase and other enzymes by the papillae in the groove, coincides with this change. In B. prionotes peak esterase production was recorded at 40 hours after flower opening (Collins and Spice 1986). In most angiospenns, secretion continues for several days after flower opening (Mattson er aI.1974). Pollination Curent knowledge of pollinating fauna associated with the Proteaceae is far from complete. Data suggest that birds such as honeyeaters are common visitors and a variety of small rodents and marsupials are also known to forage on some Banksia species. Invertebrate visitors such as bees and wasps have been identified as visitors, and in addition small arthropods have been associated with some species, such as B. prionotes (Collins and Rebelo 1987). Vertebrate visitors vary in size and foraging behaviour. For example, honeyeaters often perch on the distal part of the inflorescence and lean forward and downward inserting their bills into recently opened flowers. The forehead a¡d throat of the bird rub against the pollen presenters and stigmas (Collins and Rebelo 1987)- The success of particular animals as pollinators depends on their ability to deposit pollen on the stigmas of receptive flowers. Some species o1. Banksia such as B. bauerí and B. attenwüa attract a variety of pollinators (Collins and Rebelo 1987). Preferential foraging by honeyeaters has been reported for some species, where partly opened inflorescences maybe preferred regardless of the relative abundance of other inflorescences. The l4 Literatrue review likelihood of effective pollen transfer is also increased by birds foraging exclusively within 1-2 cm of the boundary be¡ween opened and unopened flowers (Collins and Spice 1986). V/ith hermaphrodite plants, both selfing and outcrossing are theoretically possible. The foraging behaviour of many visitors can facilitate transfer of pollen from freshly opened flowers to older receptive stigmas on the same inflorescence. However, it is likely that birds and mammals promote outcrossing to a gleater extent than insects (Collins and Spice 1986), as the former generally can move greater distances. Consequences of Dffirent Reproductive Modes. The consequence of self fertilisation is important for our understanding of breeding systems. In the heterozygous diploid, a dominant allele often 'shelters' a recessive allele which may be deleterious in the homozygous state. Self fertilisation quickly results in the accumulation of deleterious recessives. Further selfing results in the production of uniform lines differing in various vegetaúve and reproductive characters, with weakening and sterility of the plants. If plants from different lines are crossed then hybrid vigour or heterosis may be demonstrated. However, plants that are closely related and a¡ise from the same original parental stock do not give heterotic plants. The long term disadvantages of inbreeding demonstrate the advantages of the outbreeding mode of reproduction, which leads to high levels of va¡iation. The role of incompatibility is very important in considering breeding within populations. V/hile some fruits and seeds may be dispersed long distances, somo may fall near the parents and form family groups made up of genetically closely related plants. Crossing between close relatives may lead to inbreeding depression, and incompatibility mechanisms will rostrict matings between the closely related individuals. l5 Liæran¡rereview Generally, obligate outbreeding appears to have advantages, but there may be costs such also has as uncertain environmental factors influencing cross pollination. Apomixis advantages and disadvantages including the propagation of undesirable types' An analysis of the three modes of reproduction shows that each has advantages and ovrl.^ producenv/ell d.isadvantages. It would appear that there ars advantages in the short term to adapted genotypes, but the lineage unable to produce va¡iation is disadvantaged in competition with those capable of change. Lack of variation may not allow a species to is withstand selection pressures such as changes in climate and seasonal differences. It therefore not surprising that flowering plants have several modes of reproducúon' Self incompatibiliry (sr) Self incompatibilitylilthe inherited capacity of the flower to reject its own pollen' The degree of self incompatibility in Banksia va¡ies among species (Carpenter 1978' Collins and Spice 1986; Salkin 1987). Some show self compatibility following hand pollination, but natural levels of selfed seed set are largely unknown. Biochemical and genetic estimates from isozyme analysis of seeds of two species of. Banksía showed complete outcrossing (Carthew et aI. t988). The high rates impty that there is good mixing of pollen in the populations studied. Pollen gene frequencies were relatively uniform in distribution over plants suggesting that pollen is widely dispersed within the population. Results at different has sites were similar despite differences in the level of seed set. In a number of species it been been reported that outcrossing is essential for seed set. Thus, some species have reported to be self incompatible, for example B. menzíesiÍ (Ramsey and Vaughton 1991) andB. prionotes (Collins and Spice 1986), but some such as B. spinulosø (Salkin 1987) have been reported to be self compatible to some extent. Some other members of the proteaceae show varying levels of self compatibility, for example Macadamia (Sedgley 1983). More research is needed however, on breeding systems within the family. IÓ Literature review Incompatibility types can be divided into two groups, sporophytic and gametophytic. In sporophytic incompatibility, action of the pollen is determined by the genotype of the plant producing the pollen. In gametophytic incompatibility, the genotype of the individual microspore determines is action. The system of incompatibility can further be classified into two groups: (1) Polyallelic series at one or two loci (rarely several) in gametophytic and sporophytic systems where the flowers are homomorphic. (2) Two alleles per locus at one or several loci, typically sporophytic and the flowers are heterostylous. The sporophytic system has a dry stigma has so the pollen has direct contact with the papillae. In contrast, the gametophytlc system a wet stigma surface on which there is no direct pollen stigma contact, since germination takes place in a fluid medium. Incompatibility expression can be observed in three sites in the flower: the stigma, style, and the ovary. Depending on the system involved these sites can be used to infer essential points of the incompatibility system in the species. Stígma. The stigma reaction is characteristic of species with tri nucleate pollen (Brewbaker 1959)' whereas those species that have binucleate pollen are more tikely to show inhibition in the style or ovary. Trinucleate pollen is more specific in growth requirements invítro and it is suggested that inhibition in the stigma is due to lack of stimulus on the stigma via the incompatibility reaction. Stigmatic incompatibility is generally restricted to sporophytic systems (tleslop-Harrison ¿t aI.l974).It is suggested that incompatibility proteins held in the exine of the pollen grain become immediately available on contact with the stigma- Substances responsible for the stigma reaction are of tapetal origin @ickinson and læwis r973). LI Liæranrrereview Stylar inhibition Inhibition of the incompatible pollen tube in the style occurs from a few to many hours after germination of the pollen. It is cha¡acteristic of self incompatible plants with binucleate pollen. Stylar incompatibility is generally restricted to gametophytic incompatibility and usually leads to swelling and bursting of the pollen tube apex in the upper region of the style. The fact that incompatible and compatible pollen tubes do not (Linskens influence each other suggests reactions occur on the surface of the pollen tube 196s). Ovary inhibition have In a number of cases the incompatibility reaction does not act until the pollen tubes reached the ovary and in some cases initiated gametic fusion. In some species where incompatibility occurs in the ovary, the styles are hollow and do not provide pollen tubes with the conract to the stylar tissue necessary for inhibition (Brewbaker 1959). Ev oI urt o n of s eU i nc omp atibi líry As outbreeding operates in both gymnospenns and angiosperms, it appears that self incompatibility, which is restricted to the angiosperms, must have coincided with the expansion of the phylum and therefore occurred as a primitive character in the mid- Cre\c,¿gor.{3 ääåäöÈperiod (Whitehouse 1950). The rise of a very efficient outbreeding system, such as self incompatibility, is likely to have occurred in conjunction with the evolution (Grant of specialised pollinating insects at a very early stage in angiosperm evolution Ig¡g).Another line of evidence that self incompatibility is a primitive character is that its occrurence is restricted to centres of origin of species (Emst 1953). 1E Literature review T læ s elf íncomp atíbílíty mechanism proteins All types of incompatibility appear to proceed as follows: (1) Accumulation of S in the pollen walls and in the inhibition sites in the pistil. (2) Recognition, after incompatible pollination, of identical S proteins in pollen and pistil. (3) Failure of the pollen grain to establish or maintain contacr with the stigma and the formation of callose below the pollen within the papillae of the stigma, or in the case of stylar incompatibility the breakdown of the callose-rich inner wall of the tube apex and ttre accumulation of wall ìL'nhibition particles in the tube cytoplasm. (4) +niUige*after this phase or through the process itself of the pollen grains or tubes. I nter sp e c ifi c I nc o mP atibilitY Interspecific incompatibility is defrned as any post pollination process preventing formation of hybrid zygotes. This may occur through the absence of pollen germination or via abnormal behaviour of pollen tubes. The phenomenon prevents gene flow between species, in contrast to self incompatibility, which restricts inbreeding, and establishes upper limits to outbreeding and panmixis. Interspecific incompatibility thus contributes to the isolation of populations and favours speciation and the gradual increase of polymorphism within the genus and within the family. It acts as a breeding barrier between sympatric species, but also participates in the rejection of foreign germplasm migrating as pollen grains from allopatric populations, and in the isolation of invaders introduced as seed. Although less studied and far less understood than self incompatibility it displays characteristics in common with the system of self rejection. Interspecific incompatibility is usually stylar in the genera where SI is expressed in the style and stigmatic in families characterised by stigmatic SI. An important feature of interspecific incompatibility is that it usually occurs unilaterally and often prevents self incompatible species crossing with self compatible species (Harrison and Darby 1955). T9 Literature review Unilateral incompatibility also occurs between self compatible species and also in populations of self incompatible plants. Interspecifrc incompatibility occurs throughout the angiosperms in a wide range of families. It is the thesis of Hogenboom (1973 , 1975) that interspecific incompatibility, termed incongruity by this author, is a process completely distinct from SI with no association with the S locus. Arccord.ing to Hogenboom the non functioning of the pollen pistil relationship may be due one of two separate mechanisms. One is incompatibility' a mechanism that prevents or disturbs the functioning of the relationship and regulates inbreeding and outbreeding at the intraspecific level. The second is incongruity, the incompleteness of the relationship. In the d.ifferent species the matching of the genetic systems is not complete. In contrast to incompatibility, which is an evolutionary solution through a precise reaction to the negative effects of inbreeding, incongruity is a by product of evolutionary divergence, which may affect any part of the relationship between pistil and pollen and can be very different in each case. Hogenboom justifies the hypothesis on the basis of anatomical, physiologicat and genetic studies which suggest basic differences berween the two processes of incongruity and incompatibility- With regard to anatomical and physiological differences however, the majority of observations show that SI and interspecific incompatibility affect the same sites. The plant breeder often wishes to break down these incompatibility barriers to transfer germplasm from wild relatives into cultivated species. Reproductive Ecology Flowering plants generally produce more seeds than a¡e needed for individual replacement, because of losses during germination and plant establishment. However, low seed set is a common feature of many angiosperms, and may be a result of pollination failure, seed mortality, predatation, limited resources or genetic constraints. In some species of Banftsia, such as B. larici,?d, resource allocation affects final seed oulput zt) Literanuereview (Stock et aI. L989). It appears that low seed set in Banksía is the result of selection for obligate outcrossing and high seed mineral composition. This maintains genetic diversity of a restricted number of high quality seeds equipped for establishment in the low nunient soils of their natural habiøt (Stock et al. 1989). The seeds of Banlcsia carry large reserves of oil, protein and essential mineral nutrients, especially phosphorus, and trace elements (Kuo er at. L982),and seedlings aro able to grow rapidly in soils deficient in phosphorus and nitrogen. Banksia seeds are unusually rich in proteins which are mostly associated with distinct protein bodies containing globoid inclusions, enriched with mineral elements such as phosphorus, calcium, sulphur and magnesium (K;ulo et al. 1982)' The development of highly nutritious seeds also provides predators a good quality food source in a poor environment. Predators such as birds and insects reduce seed production of Banksia The valves of the follicles in which the seeds develop are held together by a resinous substance that is destroyed by heat. The structure of valves is quite complex, consisting of three layers of sclereids in various orientations. As the follicles dry, stress increases in the valves, which is relieved when the resin Iayer melts, causing the valves to open to allow release of the seed (Waldrop 1983). However, heating follicles does not always release ttre seed. Cowling and Lamont (1986) found that wet-dry cycles increased the rate of seed release. It was suggested that the separator in the follicle is hygroscopic, and with alternate wetting and drying it acts as a lever easing the seeds out. Fire is also an important component of the life cycle of banksias . Banksia species differ in their response to fire. Obligate seeders are killed by fire and regenerate from seedlings, whereas resprouters reglow from epicormic buds beneath the bark, or from underground lignotubers (Zartmit and V/estoby 1988). While many plants rely on fue to promote germination of seedlings, or regrowth from lignotubers, the frequency of fire is critical to the survival of the species. If fire occurs after too short an interval, plants may be destroyed before they have flowered and set seed. Ultimately, plants which are obligate 2T Liæraturereview seeders may become extinct with too frequent burning. Studies on the behaviour of banksias in fire has been conducted with species such as B. hookeriana (Lamont 1985), with a view to understanding regeneration limitations, and to recommend burning intervals. The seed bank at any given time will depend on factors such as growing conditions, predation and the last fire interval. Another factor governing survival after fue is the season in which the fîre occurs. Summer/autumn bums give better establishment of seedlings than spring burns, as the weather conditions are more favourable for germination. Seed released in spring will not germinate until the following autumn, and during this time it may be eaten or lose viability. Seed mortality is very high in the fust two years after germination with only a few seedlings surviving to adulthood- R epro ductiv e is olation The mechanisms of reproductive isolation can be temporal, ecological, or physiological and may exert their effect pre or post zygotically. Differences in flowering time can be an effective breeding barier and although there may be times when flowering overlaps, the likelihood of producing hybrids is low. Ecological isolating mechanisms such as the absence of common pollinators, and niche conditions select against intermediate hybrid plants. Physiological isolating mechanisms can be no resultant seed set, or reduced viability of seedlings. While the concept of increasing taxonomic distance and increasing genetic disharmony is widely accepted, the amount that this contributes to the extent of hybridisation, in relation to other factors such as incompatibility is not known. Natural Hybridisation Presumed natural hybrids occur in wild populations, the most common being B. robur xB. oblongíþIia (George 1987). Other putative natural hybrids a¡e shown in Table 2.2 (Taylor and Hopper 1988). Confirmation of their status relies on a number of criteria including the presence of both parental species in the immediate area, morphological characteristics of the 22 Lieran¡rereview hybrid intermediare between both parental types, a degree of pollen sterility, and use of techniques such as protein electrophoresis and DNA based methods to establish hybridity. Most Banksía hybrids have yet to be confirmed through pollen, protein, and DNA techniques. The hybrid between B. prionotes andB. lindelyara is significant since the two species are not closely related, and are classif,red in separate series (George 1981). Only a single tree exists, the fruit is like B.Iindelyanabutthe leaves a¡e intermediate between the two parental species. A single hybrid of B. oblongfolia and ^8. integriþlia was found to have the leaf shape of B. integriþIia, except for a few blunt teeth, and a white undersurface (8. íntegriþlia),but firm texture and rusty hairs on the midrib (8. oblongfolia).The flowers had the more silky appearance of B. integríþlic rather than B. oblongfolía. Another presumed hybrid B. saxicolaxB. marginatahas leaves similar to B. saxicolabut the fruiting cones retained the old flowers and appeared intermediate between the two species. A single old tree presumed to be a B. hookeríana and B. attenwfia hybrid had more than 1000 flower spikes (Taytor and Hopper 1988), but they were small in size and the number of flowers was very low. The flowers were also malformed and never opened properþ. The plant was infertile and did not produce seed. Leaves were intermediate and old leaves were retained, as n B. hookeriana. Table 2.2 Recorded hvbrid banksias Female parent MaIe parent B. ericiþlia x B. spínulosa B. marginata x B. conferta B.marginata x B. integrifolia B. aemula x B. serrata B. paludosa x B. integriþlia B. hookeriana x B. prionotes B. paludosa x B.marginata B. saxícola x B.margínata B. integrifolia x B. oblongiþlia B. ericíþlia x B. spinulosa B. prionotes x B.lindleyana B. prionotes x B. hookeríana B. hookeriana x B. attenu.ata B. hookeríana x, B. menziesíi 23 Literanrereview Banl B. hookeriana.x B.$r¿neì¿;and B. hookerianax B. menzí¿sii. (George 1987, Sedgley er al. I99I). Hybrídfitness çuch cnr EL¿cø\sdts In many cases,^natural putative hybrids are Fl progeny, with low variability and morphological segregation of parental characters in the seedlings (Ashton and Sandiford 1988). These hybrids may arise due to chance such as following fue. However, their contribution to the gene pool and thus to gene flow between the species will depend on reproductive output and the adaptiveness of their progeny. The three phases of hybridisation described by Drake (1980) are: F1 plant establishment, plant fertility, and evoluúonary potential. Some hybrids are reported to be less fertile than average compared to the parents, some are intermediate, and some are better than the parental species and can expand into the parent species' habitats. If the Fl plants are fertile, then the next most important criterion is interaction with the local environment, which determines to what extent the Fl and subsequent generations survive. Hybridisation has also been suggested as a mechanism of species migration (Potts and Jackson 1989, Potts and Reid 1988). In some cases the hybrids may colonise local habitats better than either parent, or provide a link for genetic combinations with a nearby species via long distance pollen dispersal (Potts and Reid 1983). 24 Literannereview P att e r ns of hyb ri di s ati o n In natural conditions the frequency of hybridisation depends on the co-occunence of species pairs, synchronous flowering and the presence of pollinating agents. The problems of geographic isolation are overcome in mixed species plantations and flowering times can vary with individuals. Manipulative hybridisation also increases the range of pollinations possible. Another factor determining the success of hybridisation is the extent of reproductive isolation. This can be assessed on the basis of species pair relationships. Chromosome Studies Karyological evolution within the Proteaceae has been discussed by Smith-White (1959), Ramsey (1963), Johnson and Briggs (1963) and Venkata Rao (1971). Banksia is reported to have 28 chromosomes (n=14) in all eight species tested, B. aspleniþIia, B. ericíþlia, B. integriþIia, B. Iartfolia, B. marginata, B. robur, B. serrata and B. spinulosa. The chromosomes are very small, as is common in perennial tree crops. No natural polyploids or other chromosome abnormalities of Banl sister ta,xa to Banksia. in the tribe Banksieae, also has 28 (n=14) chromosomes. Propagation In natural stands banksias reproduce almost exclusively through sexual reproduction. Some species have lignotubers which regenerate after fue, but this does not lead to the formation of new individuals. The majority of commercial Bankslø plantations are currently established using seedlings. Seeds germinate readily, with only a few species needing cool conditions to break dormancy (George 1987), and seed remains viable for many ye¿ìrs. Once seedlings are established damping-off is a potential problem in the f,ust few months. lVhen selecting particular forms for cultivation, seed propagation can lead to undesirable 25 Literatu¡e review variation in the progeny. For this reason, vegetative propagation is a goal for the production of superior plants. A number of species can be grown from cuttings, but success has been limited with the cut flower species. In general, those with slender glabrous stems such as B. spinulosa, B. eríciþlia, andB. íntegriþlía will form roots most readily (George 1987). Cuttings usually form roots without hormone treatment, and the plants must be hardened and allowed to develop in the por before planting out. Grafting of superior plants onto seedling rootstocks is another potential method, although problems wittr graft incompatibility and low success rates render it difficult. Grafting has been attempted to overcome Phytophthora susceptibility by grafting onto a resistant stock such as B. spinulosa or B. eríciþlia (Dixon et aL 1985). There has been some success, for example, B. speciosa onto B. íntegriþlia- Australian plants are increasingly propagated by tissue culture, which can increase plant numbers when hard to multipty by other means (Taji and V/illiams 1990). Tissue culture has also been used to facilitate and supplement conventional breeding programs, for example, by anther culture, embryo rescue and somatic hybridisation. Tissue culture has some advantages over conventional techniques of propagation including rapid multþlication, delivery of plants at any time of year, propagation of plants that are difficult to root by cuttings, or have a low seed set, and disease free plants. Tissue culture is more expensive than seed or conventional vegetative propagation, which limits its application. Australia has an expanding micrepropagation industry due to export markets and increased activity in producing new horticultural and floricultural va¡ieties. Research is currently underway to develop tissue culture techniques for Banlæia. C ommer c ial I mp or tanc e and I mpr ov ement o/ B anks ia Ornamental species of the family Proteaceae are now well established in the international and domestic cut flower industries. Widely accepted genera include Banksia, Protea, 26 Literanrereview Leucospermum andLeucadendron.The large showy blooms make them ideal as standa¡d flowers in arrangements, however, this focuses attention on the quality of the bloom. Therefore, the challenge to plant breeders is to improve bloom quality and other desirable characters such as yield, production time, novelty and reliability. The long production period and location in the southern hemisphere give Australia an opportunity to become a supplie¡ of cut flowers, foliage and potted plants to world markets. Ausralia, even with a small penetration to those markets could earn significant export income. Exporters and wholesalers buy fresh flowers from growers and air freight to destinations such as the USA, the Netherlands, Japan and Switzerland. Large quantities of flowers and foliage of banksias and other Australian species are exported. The future for Banl Banlæia flowers last well when cut, and can be dried for use in floral affangements. The decorative woody fruits can also be used by floral arrangers, while spent flower heads and fruits can be used in craft. Banksias are important for amenity horticulture, as they vary considerably in habit from prostrate bushes to tall trees up to 15 metres. Flower colour covers a wide spectrum including red, orange, brown, yellow, cream, purple and pale mauve. In some species, such as B . coccinea, a range of colours from deep red to orange occurs. Banksias are popular for their bird attracting qualities when nectar flow is at its peak, and honeyeaters are frequent visitors. At night pigmy possums feed on the nectar of some species, and insect eating birds also frnd ample food as the flowers attract insects. Banksias are grown commercially for cut flowers in South Africa, Israel, USA (Hawaii and California) and Australia. From a genus with over 75 taxa there are approximately eight species, of which B. coccine¿ is the most popula¡, that are currently exploited. Further species have potential to be used in commercial horticulture. Resea¡ch is essential to understanding the biology of banksias and to develop strict selecúon criteria for crop '¿'l Literature review improvement. Developing selection criteria is one of the most important stages of a breeding progmm, and for most proteaceous crops the criteria are yield, quality, and Phytophthora tolerance. Phytophthora is a devastating disease which affects many members.of the family Proteaceae and can be a limitation to the cultivation of plants for cut flower production. In addition, this disease can wipe out existing populations and decrease natural resources. Yietd is the major factor determining the economic viability of a ptanting. Presently, plantations are established via seed and show great va¡iation in yield. Through breeding and selection, yield gain can be achieved and vegetative propagation can be used to capture the yield gain, and thus increase efficiency of plantation management. Complete uniformity needs to be balanced against the advantage of mixed crop resistance to environmental pressures, which is especiaily important in long term crops such as banksias and a mix of improved cultiva¡s is an urgent aim. Quality is very important in marketing and strict quality standa¡ds, are being adopted in the industry over a range of crops. Important criteria include: long straight stems, terminal blooms with no interfering foliage, even bloom with minimal aborted florets, attractive floret colour, complementary foliage with attractive colour and shape, and no pest or disease damage. The South African and American selection and breeding programs for Proteaceae species have resulted in many cultivars that have established the industry (Parvin 1981, Brits et aI. 1983). These programs have focused on the African genera, and recentþ there has been a progïam established for the Australian genus Banksía (Sedgley et aI.I99I). The breeding program focuses on the species B. coccínea, B.menziesii, B. hookeriana, andB. prionotes, and three cultivars have been registered already. The South African and American programs have used interspecific hybridisation as the basis for cultivar development (Parvin 1981, Brits 1985).The Banlcsiø program (Sedgley et aI. I99t) also aims to use interspecific hybridisation within the genus to produce novel cultivars and combine characters from other species. Hybridisation methods have been developed for Banksia based on knowledge of the breeding biotogy of the genus (Fuss and Sedgley 1991). In addition, research into the structure of the pollen presenter and the stigmatic 28 Literanrereview gloove has shown that there is more than one type of pollen presenter, and the location of the stigmatic groove is d.ifferent depending on the species (Sedgley et al- t993). This is important in interspecific hybridisation to make sure the target is clearly identif,red for deposition of interspecific pollen into the stigmatic groove. The continued interest in Ausralian species, particularly those from family Proteaceae' for cut flowers provides a challenge to plant breeders to supply new cultivars. Therefore, continued selection and cultivation of new variants and species combinations is essential to expand crurent markets and maintain high quality products. Importantly, conservation of natural populations of indigenous plants will ensure there is sufficient resource to tap for further crop deveþment and improvement in the future. I nt er sp e c ifi c Hy bridi s ati o n As in many other crops, the genetic base in ornamentals can be very n¿urow. There is clearly scope for broadening the genetic base by interspecific hybridisation, both to introduce novel morphological and physiological characters and to improve specifrc traits for horticultural performance. There are few commercially improved crops which have not been improved by deliberate interspecifîc hybridisation (I-angton 1987). The advantages of combining two genotypes are many fold. It can transfer genes for disease resistance, plant or fruit quality, or increases in yield. The success of hybridisation can also give information on the relationships of the species. Hybridisation can be effective between strains, between species and sometimes between genera, the success being dependent on the degree of relatedness (Carr et aI. 1988). Wide hybridisation may be unsuccessful due to the faiture to produce Fl progeny or in terms of hybrid weakness or hybrid breakdown. The technique for artificial hybridisation is very similar to that used in manipulative intraspecific pollination. The Banksia flower is hermaphrodite and the pollen must be removed before application of the experimental pollen. Pollen is removed by passing a 29 Literature review looped pipe cleaner over the pollen presenter region of the style (Fuss and Sedgley 1991). It has been shown that banksias are protandrous so pollination cannot proceed until the stigma becomes receprive. Experiments investigating seed set and pistil age and condition have shown that maximum receptivity is related to groove opening and stigmatic secretion and occurs approximately three days after flowering in B. coccineø (Fuss and Sedgley 1991). The exact timing depends on the species and the longevity of the flower. In other genera pollinations before the stigma becomes receptive are mostly unsuccessful (Griffin and Hand Ig7g, Sedgley and Smith 1989). Fresh pollen may be applied by touching the pollen laden presenter region of the species of interest to the stigmatic gloove. Flowers are bagged prior to experimentation to avoid pollen contamination. The use of interspecific pollen presents an additional problem in the availability of fresh pollen. If non-synchronously flowering species a¡e used then the pollen must be stored. Banksiapollen has not been tested in storage and in virro methods of germination have not been reported. Viability of fresh Banksia pollen has been estimated using the fluorescein diacetate stain (FDA) (Ramsey and Vaughton 1991, Collins and Spice 1986, Ramsey 1986), but attempts to germinate pollen in sucrose and nutrient media were unsuccessful (Prakash 1986). Pollen longevity has been estimated using pollen of different ages on fresh pistils of B. spinulosa that have been stained with acetocarmine (Ramsey and Vaughton 1991). It was noted that when flowers opened pollen viabitity was high, but quickly decreased. Considerable variation between inflorescences was also noted. AIt er nntív e metho ds of hyb ridis ation Ba¡riers to effective fertilisation occur between many potential crosses and to overcome these barriers different techniques are used to bypass or reduce the effect. As the site of rejection in many crosses is the stigma or style, many methods facilitate the entry of the pollen into the ovary. Compatible pollen has been used to stimulate the post pollination fesponse. This pollen can be used fresh or killed, or as pollen extract. The function of the 3U Literature review pollen is thought to provide the appropriate recog"ölï1,,ìåOttances missing from the foreign species, and hybrid seed has been obtained in Pdbusing this technique (Knox er al.1972). Another merhod is the use of solvents on the stigma surface. This is thought to interfere with the maternal recognition system, and hybrids of. Eucalypr¿¿s have been obtained using this method (Pryor and Willing ß7Ð. Temperature can also affect the degree of inhibition of pollen tube growth in styles (Franken et a\.7988). Some techniques involve reduction of the distance that the pollen tube has to grow. This is done by using the shortest styled species as the female, or by amputation of the style and pollination of the stump (Raff 1983). In some interspecific crosses embryos are formed but later abort, due to non function of the endosperm or interaction with the maternal genotype. Embryo rescue is used to overcome this, and can also be used for rapid multiplication of clones and acceleration of germination. Another approach is to bypass sexual reproduction via somatic hybridisation. Protoplasts of two species are fused, screened for hybrid cells, and the plants regenerated in vítro (Vasil and Vasil 1980). Bridging species can be used to transfer genes from one to another. This method is not popular due to the long time required and to undesirable characters passing from the bridging species. Hybríd Verification The most common method ussd for the verification of hybrid plants is screening of seedlings for morphological characteristics intermediate between the putative parental species. Hybrids have been identified using cotyledon shape (Kapoor and Sharma 1984), phylotaxy (Brooker 1979), and leaf shape (Ruggeri 1959). Seedling morphology has been used in hybrid verification in other genera, such as Rhododendron (Rouse et al. 1935). This technique works well when morphological markers are present at the seedling stage. Isozyme analysis is another technique commonly used. Species genotypes can be established and F1 hybrids are identified by their combination of parental 31 Literanuereview isozymes (Chapa:ro et al.lg87, Parfitt et al. L985). Recently, a technique based on the polymerase chain reaction (PCR), called random amplif,red polymorphic DNA (RAPD) analysis has been used to verify the hybrid origin of plants (Chong et aI.1994)- Pollen Collection and Uses. V/hen collecting pollen from the field, the quantity available, season in which it can be obtained and the method of handling and storage depends on the plant species. In insect pollinated plants, the insects that visit the flowers do so primarily to collect the nectar at the base of the style, or the pollen from the dehisced anthers. Pollen in such flowers is readily located. In contrast, pollen from wind pollinated species needs to be collected before anther dehiscence which can be more diffrcult. In collecting pollen it is essential that the pollen be free of contamination and that the genetic purity of the pollen be assured. Pollen may be used in hybrid production, when specific progeny are required, and known patental sources of pollen are used in the breeding Program. In some species, pollen from plants of early va¡ieties may need to be stored to facilitate pollination of late maturing varieties, or pollen from species with different flowering times may need to be stored to allow the production of interspecific hybrids. F actor s C ontrolling P ollen Availability Genetic controls The most important factors determining the time of pollen development and anther d.ehiscence are inherited. Genotypic variation and blocks to hybridisation are often established when pollen dispersal does not coincide with stigma receptivity of plants within pollinating distance (Stanley and Kirby 1973). Time of pollen shedding can be 32 Liæranuereview related to geographical conditions and anther dehiscence depends in part, on meteorological cond.itions which can change day to day. In most plants the pollen quality as measured by pollen germination remains about the same throughout the period of dehiscence. External controls Temperature and moisture are the primary elements that affect pollen development in mature plants. If anther dehiscence is early in one location then it is possible to find similar genotypes at a higher elevation or colder location where the pollen has not yet been shed. Va¡iations in temperature and moisture can also shorten the interval of pollen dispersion from a plant. Low or high temperaturos during pollen development can adversely affect the quantity and quality of pollen, and moisture and nutrient status of the plant can affect pollen viability and abundance. Collecting úmes can affect the quality of pollen, as pollen collected in the morning is often more viable than that collected in the afternoon. Radiation can have a negative influence on development, as exposure of pollen to ultra violet light can injure some pollens, so pollen is best dried in darkness or diffuse light. Chemicals applied to plants for insect control or defoliation can adversely affect pollen quality, and plants exposed to fungicides, herbicides, or pesticides applied during flower development will often yield damaged pollen. Awa¡eness of the factors that influence pollen germination from field gïown plants can help the plant breeder select plants and recognise sources of deviation in pollen behaviou¡. Pollen Storage The ea¡liest reporrs of storage and transport of pollen ffir6oo¡ 1000 B.C. when date pollen was transported by hand to pollinate female flowers (Wodehouse 1935). The 33 Literature review maintenance of pollen germination capacity depends on the conditions of storage- Critical factors include: relative humidity, temperature and the atmosphere surrounding the pollen. Relative humidiry ß.H.) The humidiry of ttre air during storage affects longevity. Most species retain their viability best at low relative humidity. Various reports indicate that pollen longevity, in general, increases with reduction in RH to about 6Vo ùtnng storage (Stanley and Linskens 1974, Akihama and Omura 1986). Many species lose viability at very high or very low RH. However, low moisture levels must be avoided in the storage of certain pollen, such as Tulipawhere the water content cannot be brought below a critical level of 407o (Stanley and Linskens 1974). Frequent fluctuation of RH during storage causes quick loss in viability, as pollen apparently cannot withstand variations which induce alternate phases of low and high metabolic activity. For many pollens pre drying is essential before storage. t Temperamre Another important factor that affects pollen viability during storage is temperature, and many pollens can be stored successfully at temperatures below zero (Sedgley and Grifñn 1989). It is assumed that in pollen stored at extremely low temperatures, such as liquid nitrogen -196 0c, the metabolic activity is zero with potential storage for an unlimited time (Sedgley and Griffin 1989). Gas atmosphere and oxygenpressure Correct atmosphere, such as increased levels of carbon dioxide when pollen is stored over dry ice, can Folong viability, however, storage in pure oxygon can shorten viability '34 Liæraturereview (Stanley andlinskens 1974). Reduction of the partial pressure of oxygen can prolong the viabiliry of some pollen. When pre dried under reduced pressure pollen can retain high germination. In comparison some pollens die when placed under reduced pressure (Stanley and Linskens 1974). Causes of decreasedvíabilíty during storage Little is known about the influence of change during storage on the genetic and physiological capacity of pollen. The primary reason for decreased pollen viability in storage is probably related to enzyme activity, which decreases respiration substrates, as the mechanism by which pollen retains its viability during storage is related to intracellular rates of respiration. This is confrmed by the requirement of stored pollen for higher concentrations of sugar for normal germination than fresh pollen, and as the respiration rate decreases with age, so the boron sensitivity increases (Stanley and Linskens I974). Activity of enzymes and changes in endogenous hormones can also decrease viability. Reduction of germination capacity under certain storage conditions can therefore be attributed to inactivation of enzymes and metabolic substrates essential for germination. Other factors may affect pollen survival in storage, such as accumulation of toxic cell compounds, oil deposits in the exine, mineral nutrition of the plant during pollen development, bacterial or viral contamination, and chemical additives from bee bodies. Víability tests After pollen is collected or removed from storage, assessment of the capacity to germinate and grow normally is important. In an in vivo assay a long time can elapse between pollination and seed set, and such a test may not be valid, since incompatibility reactions may inhibit pollen growth in the style even though the pollen can germinate normally. 35 Lite¡ah¡re review Viability tests which can determine growth potential have been developed. These include invitro assays and non germination tests. In vitro assays Most pollen viability tests germinate a small sample of pollen and observe under the microscope the percent pollen grains producing tubes after a given time. Such tests assume that optimum conditions have been established, so that germination approximates that on the plant. However, most pollen tubes stop growth before they reach the length they normally obtain in the style, and the rate of pollen tube growth is often not as rapid as invivo. Another assumption is that the sample of pollen tested is typical of the source. This assumption is permissible if care is taken to ensure that the pollen is mixed before drawing the sample. Plant to plant variation must be guarded against in drawing a sample from bulk pollen for viability testing. Non germination assays Tests which stain pollen grains with a coloured dye are often used as indices of viability. In such tests the chemicals a¡e absorbed into specif,rc cell constituents present in the mature pollen grain. Examples of such stains include acetocarmine, aniline blue, and poøssium iodide. Most stains are not suffrcientiy accurate when compared to germination tests, and can give only crude estimates of viability. Viable pollen grains contain functioning enzymes and certain dyes on reduction change from a colourless compound to a coloured form. Active enzymes in the pollen grain are usually necessary to catalyse the colour change of the dye. Examples of these dyes are TTC (2,3,5-triphenyl tetrazolium chloride), or a fluorescent reaction with fluorescein diacetate GDA). 36 Liæratr¡¡ereview Recording dnø A false germination count can result under cefiain conditions. In certain germination conditions a minimum concentration of pollen grains must be placed in a given volume of solution if maximum germination is to be attained. Conversely, high pollen concentrations can inhibit germination. Care must be taken in the viewing and counting of the pollen grains. Pollen spread over agar or a slide can give slight distortion of viability counts, if the grains are not evenly dispersed. Growth inhibiting or stimulation chemicals must also be recognised as potential sources of error. Inhibition products such as organic acids can diffuse out of the pollen grains and accumulate on the medium. Water source, pH and nutrient factors can also affect the results. In addition, the rehydration of stored pollen before viabitity testing is important for maximum germination. Comparison of viabíIity tests It should be noted that fair to good seed formation may occur even though ttre results of a invitro pollen germination assay indicate low viability. The same problem holds true for results of non germination assays. If the test indicates 4OVo or better the pollen may be quite satisfactory for use in freld pollination (Stanley and Linskens 1974).Inert materials are often used to dilute pollen to lower the percent viable grains in the sample. The capacity to form seed and seedlings is the ultimate test of pollen viability, and thus different pollen viability methods must be judged for their relative merits. The ability of pollen to grow or respond to in vivo or in vítro assay is dependent on the inherent chemistry of the pollen, which can provide an understanding of the metabolic factors faciliøting growth and seed formation. 37 Literaû¡re review Molecular Techniques to Study Relationships ín Plants There are a number of molecular techniques available to the plant biologist and it is impoftant to select the one most suitable for the problem. One of the most important factors is the degree of relationship among the organisms being studied. If the organisms are closely related, a technique which detects highly variable regions of the genome is most useful. V/hen working with distantly related organisms more slowly evolving regions of the genome are needed. Different molecula¡ ma¡kers provide varying amounts of information about the actual va¡iation at the nucleotide level and can have varying levels of usefulness. Some techniques provide small numbers of highly informative characters, while others have large numbers of less informative characters. The breeding system of the plant is also important in the choice of the marker, since this affects the levels of variation expected. Types of markers used to study plant relationships include restriction fragment length polymorphisms (RFLP) (Byrne et al. 1994), random amplified polymoqphic DNA (RAPD) (Huff et al. t993, Chalmers et aI. L992' Russell ¿r al. 7993), DNA sequencing (Phitlips et at. L994, Sang et al. 1995) and techniques to srudy repetitive DNA (Rieseberg et aI.1990). There are many other techniques available, but the following are amongst the most commonly used for studies of plant species relationships. Thc polymerase chnin reaction (PCR) technique The PCR technique rests on the use of synthetic primers to prime the synthesis of a complementary strand of DNA using the thermostable polymerase enzyme obtained from a thermophilic bacterium (Taq I polymerase). Primers are used for the two complementary strands at the opposite ends of the region to be amplified. After one round of amplification, the reaction mix is denatured through heating and allowed to reanneal with excess primer, and the polymerase reaction is carried out a second time. Repeated Jö Liæran¡rereview cycling of these conditions results in a geometric increase of the sequence determined by the initial primers. Random amplifie d polymorphic DN A ( RAP D ) RAPDs were recently introduced by Welsh and McClelland (1990) concurrently with Williams et aI. (1990).They are a PCR based technique and as such require only a small amount of DNA and tissue. This is an important consideration in non destructive sampling of rare and endangered plants, or when experimental material is scarce. The technique differs from conventional PCR in that the primers a¡e shoft arbitrary sequences, in contrast to primers that are complementary to a known region. The resulta¡t bands correspond to regions in the gsnome where primers anneal to form a product able to be amplified with the amplification enzyme Ta4polymerase. The distance between the primer sites determines the size of the product. Alttrough reaction conditions can alter the number of bands produced, it is usual to optimise the reaction conditions for a given species or genus. A primer is assessed as suitable if it will amptify a scorable number of bands. Some primers amplify too few bands, and some amplify roo many, which may be uninformative. Due to the high levels of variability detected using RAPDs.,they are very useful for population studies (Huff et a|.7993, Chalmers et at.1992, Russell et a|.1993, Dawson et a\.1993, Yu and Pauls 1993), in the identifrcation of cultivars @unemann et a\.1994, Mailer, et al.1994. Schnell, et aL 1995) and hybrids (Chong et al. 1994), in mapping genomes (Carlson et al. 1991, Cu et al. Lgg2,Lodhi et a\.1992, Grattapaga\ia et at. I992),parentage determination (Welsh ¿r ø/. 1991) and estimation of gene flow (Arnold et al. I99l). RAPDs are generally considered suitable below the species level. However, they have been used in taxonomic studies and have confirmed existing classifications based on more traditional methods, such as morphology, cytology and enzyme electrophoresis (Demeke et aI.7992,Howell et al. t994). 39 Literature review RAPDs are dominant markers, and as such cannot distinguish between a homozygote and a heterozygote; however the large number of polymorphic markers produced offsets this disadvantage. A major advantage of RAPDs is that no prior sequence information is required,. and the technique can be applied to any species where DNA of reasonable quality can be obtained. RAPD techniques are quick, relatively cheap and can process large sample sizes. In addition, RAPDs do not require radioactive labelling, a potential hazardin some other methods, and automated machines are available. Each RAPD fragment produced is separated using gel electrophoresis, and two bands which co-migrate are generally assumed to be homologous. However, it is possible that pairs of priming sites will occur in the same orientation separated by approximately the same distance more than once in the genome. Hence bands that co-migrate may not be homologous (Bachmann l994,Lynch and Milligan 1993, Stammers et aI. 1995). To check this, southern blot analysis of RAPD bands have produced varied results. In some cases the co-migrating bands are all homologous (Wilikie et al.1993), and in others some are homologous, and some are repetitive DNA (Williams et al.1990). Ideally each data set needs to be tested to confirm that co-migrating bands a¡e homologous, but this is usually not feasible given the scope of many projects. Because PCR is such a sensitive technique, some RAPD bands may arise due to contamination. In general, this can be avoided by using optimised methods and ca¡eful technique. Many workers have shown that reproducible results can be obtained if specific conditions are used and care is taken to avoid any alteration of the conditions (Virk et ø/. 199s). Debate continues as to which is the best method of analysis for RAPDs. Analysis for systematic or classification purposos can be done using similarity coefficients or methods such as parsimony using PAUP (Phytogenetic Analysis Using Parisomony) (Swofford 1990). Choices of coefficients and their implications for subsequent analysis should be 40 Literatu¡e review carefully considered (Jackson et at. 1989, Swofford and Olsen 1990, Wier 1991), as should earlier choices regarding sampling (Baverstock and Moritz 1991). Each RAPD data set should ideally be tested with different methods of analysis to determine thei¡ suitability, and compared to independent information. DNA sequencing Direct PCR sequencing enables rapid and precise determination of sequence identity and variation, which is useful in most aspects of molecular biology and genetics. PCR improves the ease and capacity of DNA sequencing activities by simplifying the screening, preparation and manipulation of the DNA templates. The technique is so powerful that it is one of the most utilised molecular techniques in phylogenetic studies (Hillis et al. L990). The sequence chosen is of the utmost importance as it must contain the appropriate level of variation actoss the taxa to be studied. The target sequence must also be present in all the organisms studied and be capable of alignment. Sequencing can be used to produce phylogenies of species (Taberlet et aI. L99l), genera (Phillips et a/. Lgg4, Sang er aI. 1995), orders (Frye and Hedges 1995), subfamilies (Hsiao et al. 1995), families (Olmstead and Sweere 1994) and to investigate relationships at higher taxonomic levels (McCoun 1995, Raff et al.1994). Molecular phylogenetic analysis has become easier because of PCR. A general strategy rhat can be used is to apply a general set of primers to amplify and then sequence the region directly using Taq polymerase (Ruanto and Kidd 1991). Usually the primers flank a hypervariable region and hybridise to conserved flanking regions such that comparisons can be made against a wide variety of genotypes. Introns are particularly good for this purpose, because they generally evolve faster than exons (V/olfe et aL 1989) and tend to be short in plants (tlanley and Schuler 1988, Hawkins 1988). 4T Literature review In order to use PCR amplification of the target sequence, the individuals need to be homozygous for that sequence (Hillis et aI.1990). If the sequences to be analysed are homogeneous a single sequence results, but if a mix of similar sequences is present, a direct seguence analysis will reveal ambiguous signals at the mixed positions. Cloning PCR products allows sequence variants to be separated before sequencing. Generating direct sequences from PCR products is technically more diffrcult than analysis of cloned single stranded DNA templates, as the PCRs generate complementary strands that can compete with a sequencing primer to bind the template, and contain salt and other reagonts that can affect the sequencing chemistry. These contaminants can lead to artefacts and completely failed roaction 'stops' in every lane. Sequence reactions are complex mixtures and it is often difficult to trace the precise sourco of a sequencing problem. PCR products can be sequenced by conventional radioactive methods, or by fluorescent automated DNA sequencing (Smith et aI. 1988, Voss et al. \989). The fluorescent methods are well suited for relatively high throughput studies, with direct computer entry of the final data. Manual methods rire expensive and time consuming. The radioactive merhods are generally more successful when templates of poor quality are used" while the fluorescent methods require more stringent variation of the reagents. This is particularly tme of the methods that mix four different colours in a single lane, as the chemistry of each sequence reaction must be carefully balanced. The current fluorescent DNA sequencing strategies work most favourably with oligonucleotides with dyes attached at the 5'terminus. PCR provides a simple alternative and fragments can be amplified by primers that have a universal sequencing recognition site at the 5' terminus. This site is incorporated into the fragment via one PCR primer and then commercially available fluorescent primers are used (McBride et al. t989). An advantage of the fluorescent assay is that the detection of the primary signal offers a linear response to the amount of material that is present over a wider range than radioactive methods. As a result fluorescent DNA sequencing via PCR is becoming the preferred method. Combinations 42 Literature review of PCR and DNA sequencing are now standard items in the molecula¡ biology tool kit and the r¿mge of applicaúons in molecula¡ biology is likely to continue to increase. Molecular Approaches to Plant Systematics Traditionally plant systemaústs have sought to determine relationships based on a wide range of criteria, including mo¡phological similarities at the gross, anatomical and ultrastructural levels, and similariúes with respect to secondary metabolites, isozymes and other protein systems. Recently DNA analysis has been used as a basis for biosystemaúc srudy (Palmer 1987, Ritland and Clegg 1987). Genetic relationships between plants can be estimated from the pattern of DNA sequence change. The investigator is faced with various choices regarding the level of genetic resolution appropriate for the materials under study. Common choices in plants include the chloroplast genome, (cp DNA) or components of the chloroplast genome, the nuclear ribosomal RNA genes (r RNA), or nuclear encoded, single copy genes. Another consideration is the method employed to provide direct or indirect measrues of sequence divergence. Genetic d.ivergence can be determined using restriction site changes. Restriction site analysis requires some knowledge of the physical map of the genome. Empirical studies have shown that restriction site analysis provides good resolution at or below the family level. A second method is DNA sequencing. Until recently DNA sequencing required molecular cloning of the gene or DNA fragment under study, which was subjected to a 'sequencing strategy' for the production of overlapping sequencing runs. Two related technological advances have overcome these necessities. It is now possible to synthesise oligonucleotides that can prime dideoxy sequencing in the gene of interest, based on previously determined sequence information. The second major advance is PCR. PCR allows direct amplification of the DNA fragments from heterogenous DNA samples, and thereby circumvents the cloning procedure. Application 43 Literanrre review of PCR to chloroplast genes, together with direct primer sequencing, provides an excellent method to obtain large data sets. Another interesting area of PCR application is in the amplification of DNA fragments from fossil plant and animal tissues. Fossil sequences have been obtained from Miocene- aged leaf compression fossils (17-20 million years old). The rbcL gene has been amplified and compared to other related plants (Clegg and Durbin 1990), confirming its identity. The study of ancient DNA opens up a whole new area of palaeobotanical research. In the past it was necessary to infer relationships of plants; now the extant plants can be analysed directly, eliminating some of the problems that complicate the study of plant phylogeny. Higher plant cells contain three genetic comparünents which separately transcribe DNA and translate messenger (m) RNA into polypeptides. Most plant proteins are derived from the nuclear DNA, with fewer genes in the mitochondrial or plastid DNA. The cytoplasmic genomes are much smaller than the nucleus and are present in multiple copies in each organelle. As there are many chloroplasts and mitochondria in the cells these genomes can contribute a large proportion of the total cellular DNA. Chlorpolast DNA in particular has been extensively used to study plant phylogeny. Chloroplast DNA Chloroplast DNA (cp DNA) continues to be used as a phylogenetic tool for many plant species (Hoot et al. t994. Mummenhoff and Koch 1994) and a number of reviews have been published detailing its suitability for such studies (Clegg and Durbin 1990). Chloroplast DNA occurs as a circula¡ molecule with multiple copies per organelle, and is relatively easy to isolate (Clegg and Durbin 1990). Most angiosperms transmit chloroplasts through the maternal lineage and progeny usually receive the entire chloroplast genome from one parent (Sears 1980). Thus the use of chloroplast DNA 44 Liærah¡rereview ensures that the DNA examined originates from a single parental lineage, rather than a mixture of recombined DNA as in the case of nuclear DNA. In addition, foreign DNA is not incorporated into the chloroplast genome, in contrast to nuclea¡ and mitochondrial genomes (Palmer 1987). It has been suggested that the chloroplast genome evolves at a slow rate a¡d is ideal for analysis of plant evolution. The slow rate of evolution may also be disadvantageous for some purposes, but should be appropriate for the intergeneric and subgeneric levels (Palmer et al.1988). PCR analysis of chloroplast DNA PCR amplification of variable regions of the chloroplast genome, such as introns or intergeneric spacers, displays high levels of variation, while conserved coding regions allow comparison over a wide rangs of taxa. Polymorphism of the PCR products is detected by direct sequencing of the product, or restriction endonuclease digestion of larger products. To amplify across a large number of taxa the region of interest must be free of structural rearrangement, although some insertions or deletions may alter the size of the product and this may be useful as a marker itself. Compared to traditional sequencing methods, PCR amplification allows the use of small amounts of tissue and heterogenous DNA samples and is also quicker and cheaper than other cloning methods. The primers used can be chosen for the level of variation required. Slowly evolving sequences can be used for divergent taxa, and rapidly evolving sequences can be used for innaspecific studies. Sequencing of intergeneric spacers of cp DNA can be used for phylogenetic sildy of closely related species or as intraspecific markers. Taberlet et al. (1991) designed primers to amplify regions between the trnT and trnF ,rrtenleyric genes in the large single copy region which included an infuie spacer region between the trnT and trnL,the trnL intron and another intergeneric spacer between the 45 Literan¡re review trnL3'exon and frnF. These were studied by sequencing the PCR products. The primers were tested across a wide range of taxa including green algae, bryophytes, pteridophytes, gymnosperms and angiosperms. The region is thought to be universal for plant taxonomy studies. Analysis of DNA sequence data Many techniques exist for the analysis of DNA sequence data. The simplest algorithm is the unweighted pair group method (UPGMA), which clusters based on similarity. A drawback of the analysis is that the statistical error associated with the topology of the tree is not evaluated, because it is not based on a probabilistic model. Pairs of sequences that have been separated for the same length of time may have different numbers of substitutions because the occurrence of a mutation is a random event. It is therefore desirable to estimate the statistical eror associated with the process of nucleotide Fe\se'rrtteirr substitution. FeÈnditi* (1981) developed a maximum likelihood algorithm for the esrimarion of phylogenetic topologies that is based in a probabilistic model of the substitution model and allows calculation of confidence intervals of the branch lengths. A major drawback of this method is that it is computationally difficult and a large amount of computer time is required for a moderate number of taxa and site differences. Resampling methods such as the bootstrap method are being used to quickly estimate the statistical li\¡enelervr error of phylogenetic topologies (Rüèníteià 1935). Sequence analysis can also be performed using parsimony methods such as Phylogenetic Analysis Using Parsimony (PAUP) (Swofford 1990). Conclusions In summary, there is little doubt that the introduction of PCR has resulted in a large increase in genetic knowledge of plant species, and that this will continue to grow. The ultimate DNA marker is DNA sequencing, and when this becomes cheap to obtain and 46 Literanrereview analyse large amounts of sequence data will be produced and some of the older methods will be superseded. In the mean time it seems that it will be possible to make genetic maps and screen thousands of genetic loci quickly. This will directly affect new crops that have not received much research attention. It will also allow ecological and population geneticists to add¡ess novel questions because of the high marker output and the capacity to analyse large numbers of individuals through automation. The range of molecular techniques now available for studying relationships in plants allows a technique to be chosen which is appropriate for the genetic question under investigation. 41 B . menziesü pollen viability afær storage Chapter Three Viability testing of Banksia menzicsü pollen after storage at different temperatures Abstract Ger m\ no,\ior, q\\er 1\ oçc^qe. 0C S+erageof Banksia menà.esüpollen was assessed.at20,4, -20, -80 and -196 using a semi solid medium of I7o agar, I57o sucrose, 0.017o boric acid, 0.039o calcium nitrate, 0.02Vo magnesium sulphate, 0.0L Vo potassium nitrate, and an incubation temperature of 25 oC. Germination remained constant in all treatments except room temperatue ,ãOOC, which after six months had only 257o germination. Pollen viability was assessed using fluorescein diacetate (FDA), but the results did not reflect the loss of germinability at 20 0C. Ttrere was no effect of floret position on the inflorescence on germination, but pollen viability va¡ied over the flowering peroid with ma-nimum germination mid season. Introduction Pollen storage is an important component of plant breeding programs, when the species or individuals to be hybridised do not flower at the same time, or are geographically isolated- In the genus Banksia,there is no published information on pollen collection and storage, or on in vitro conditions for successful pollen germination and tube growth- Storage conditions vary for plant species, and in most woody species low humidity and low temperature storage is optimal (Stanley and Linskens 1974, Akihama and Omura 1986). Dehydration of pollen to 3-57o moisture increases pollen longevity at low temperature, but the pollen must be re-hydrated before it will germinate (Yates and Sparks 1939). Pollen viability testing methods include staining with materials such as fluorescein diacetate (FDA), or acetocarmine, and invitro pollen germination (Sedgley 46 B. menziesü pollen viability after storage and Griffin 1989). Staining methods rely on the activity of enzymes in the pollen, which may penist after the ability of the pollen grain to effect seed set has declined. Similarly, in vitro pollen germination does not always reflect the true viability and lack of reproduc.ibility of results has been demonstrated in some genera including Acacía (Sedgley et al., L992). The ability of pollen to effect seed set is the most accurate test of viability, but time constraints dictate that indirect methods are widely used. There is considerable interest in cultivation of Banksia species for cut flower and potted plant industries. Recent work by Fuss and Sedgley (1991) on the development of hybridisation techniques for Banlcsia has stimulated the use of interspecif,rc hybridisation in banksia breeding (Sedgley et al., t994),and flexibility of the breeding progam would be greatly increased if pollen could be stored successfully. The objective of this study, was to develop pollen viability testing methods for Banksia, and to develop practical methods for long term storage. Materials and methods Pl^antmnterial Plants of B. menzi¿sii were located in collections at Happy Valley and Nangkita, South Australia (latitude 35010' S, longitude i38o34'E). Climate and soils data are presented for the two sites (Figs 3.1, 3.2; TabLe 3.1). All experimental plants were grown from seed, and ranged in age from 3 to 15 years. Experiments were conducted over the flowering season of B. menzi¿síí during the period of 1993 to 1995. Pollen collection Banksia menziesii inflorescences comprise hundreds of florets, and in preliminary experiments pollen was collected from the upper, middle and lower thirds of four 49 B- menziesü oollen viabilitv after sûorage inflorescences from the same plant. Pollen was also collected from the middle third of three inflorescences from th¡ee plants at monthly intervals during the flowering period. In subsequent experiments, pollen was collected in the middle of the season from the middle third of up to 15 inflorescences, from three plants. Pollen from all inflorescences of each plant we.re mixed together to give a pooled sample. Pollen storage Florets with dehisced anthers were removed and placed in 1.5 mL eppendorf tubes in a 0C. desiccator over silica gel (pre-equilibrated for 1 month) for 24 hours at 4 They were placed in containers with silica gel at 20,4, -20, -80 or -196 0C (liquid nitrogen). The tiquid nitrogen treatment had no silica gel since the vapour pressure is negligible. Pollen was stored for up to 6 months and tested following a period of hydration at 1007o humidity for one hour. Killed pollen, which was kept at 60 0C for 5 days, was used as a control. Fhnrescein diacetate (FDA) test Pollen was placed on a slide with two drops of FDA (1mg FDA in 1 mL acetone; added dropwise to 50 mL 107o sucrose, until mixture appears milky) (Heslop Ha:rison ¿f ¿/. 1984). The pollen was allowed to sit for 30 mins to allow penemtion of the FDA into the pollen grain before observation with a fluorescence microscope. At least 250 grains were counted in different areas of the slide, and scored for the presence of fluorescing and non-fluorescing grains. Grains which were bright and consistent in fluorescence were scored as viable, and those which were non-fluorescing, or had weak and patchy fluorescence as inviable. )U B menziesü Dollen viabilitv after storage In vitro germinotion The basal semi solid med.ium comprised IVo â.gar,0.0IVo boric acid, 0.03Vo calcium ni6ate, 0.02Vo magnesium sulphate, and0.0l7o potassium nitrate in 7 cm diameter petri dishes. Three sugar sources, sucrose, Iactose and fructose were tested at a concentration of LyVo. The best sugar source was tested at 5, 10, 15 and 20Vo.Inatbation temperatures 0C. tested for the best medium were 15,20,25,30, and 35 Pollen germination was scored using a dissecting microscope after incubation for 18 hours on the medium. At least 250 randomly selected pollen gtains in different fields of view were scored for germination, the criteron for which was a tube longer than the grain diameter. Pollen tube length was measured for 10 grains per replicate. Sndsñcal analysis All experiments were randomised complete block designs, using plants as blocks with three replicates pil treatment. Results were analysed using analysis of variance (ANOVA) with the staristical package Genstat 5 (Payne 1987). Correlation coefficients were calculated by ploning a line of best fit. Results In vitro germination For pollen collected from the middle third of the inflorescence in the middle of the flowering season, sucrose gave the highest germination, with glucose and lactose supporring poor germination (Table 3.2). There was no significant difference between 10, 15 and,20Vo sucrose in germination or tube length, but measurements were lower with 57o sucrose. A concentration of 157o sucrose was used in further experiments. Pollen germination was highest att5-25 0C with no significant difference between these 5I B. menzíesü pollen viabilitv afær storage 0C. temperatures. There was a drop in germination at 30 0C and almost none at 35 Temperature also had a significant effect on pollen tube length which decreased with increasing temperature. An incubation temperature of 25 0C was used in further experiments. Effect of posítion on the inflorescence and. time during the flowering season Invitro tests showed no significant effect of position on the inflorescence on germination percenrage or on pollen tube length (Table 3.3). There was a significant effect of time of season on FDA staining and pollen germination, with pollen at the start of the season showing lower viability than mid and late season (Table 3.3). The correlation between FDA staining and invitro germination was good with r 2 = 0.8. Pollen storage There was a significant effect of storage temperature, time and an interaction between temperaturo and time on pollen germination. All treatments retained germinability, with the exception of 20 0C (Table 3.4). The effect of storage temperature on pollen tube length was nor significant, although time in storage was. Pollen viability as measured by FDA staining showed no significant effect of temperature, with all treatments showing similar results (Table 3.4). There was a significant effect of time only. The correlation 0C coefficients for the germination and FDA tests were low, in the case of 200C and -80 and high for 4 0C, -20 0C and -196 0C (Table 3.5). Killed pollen showed zero germination andFDA staining indicated 3Vo vtabibty. 52 B. menziesü pollen viability afær storage Tabte 3.1 Soil profile of Happy Valley and Nangkita field sites. Soil classification according to Northcote (1979). Site Subsite Soil Depttr Texture pH Concl- Structure class (cm) uctivity dune sand gram Happy 60-70 Loamy 6.0 4.7 Loose Valley sand grain 70-80 Loamy 6.0 3.8 Loose sand grain B Dy5 Dune o-25 l¡arny 6.0 z.E Loose swale sand grain 25-55 Medium 5.5 ß.2 Massive claY UcZ Plain 0-15 I.oa¡ny 6.0 3.2 Loose (<57o) sand grain slope Nangkita 15-55 Sand 6.5 2.9 Loose grain 55-70 Loamy 6.0 3.8 Loose sand grain )J B. menziesü oollen viabilitv after storage Table 3.2 Optimisation of in vitro pollen germination medium for B. menziesä Variable Pollen germination 9ù Pollen tube tength (Pm) Sugar source (1,0Vo) Incubated at25oC Sucrose 6r.4 97.5 I-actose 12.0 101.3 Glucose 20.3 100.0 Probability <0.05 NS S ucrose concentrati on (Vo) Incubated at25oC 5 18.1 67.5 10 50.6 96.3 15 ó8.9 100.0 20 60.1 r2r.3 Probability <0.01 <0.001 Incubation temperature (oC) 107o sucrose medium 15 57.9 r37.5 20 56.4 1t6.3 25 57.0 100.0 30 13.6 77.5 35 0.5 28.8 hobability <0.001 <0.05 54 B Dollen viabilitv after storage Table 3.3 Effect of position on the inflorescence and time during the flowering season on viability of pollen of B. menziesü. Variable Pollen germination Pollen tube length Fluorescence (Vo) Grm) (Vo) Posiúon on inflorescence Upper third 72.2 88.8 Mddle third 77.6 9r.3 Lower third 71.8 96.3 Probability ns NS Time during season June 63.4 115.0 52.2 July 75.8 96.3 82.9 August 70.2 90.0 83.1 Probability <0.001 <0.05 <0.001 Not tested 55 B oollen viabilitv afær storage Table 3.4 Effect of storage temperature and time on viability of pollen of B menziesü. Storage cond.itions Pollen germination Pollen tube length Fluorescence (Vo\ (pm) (vo) 20 0c 0 months 84.6 100.0 85.4 3 months 53.1 87.5 62.0 6 months 26.3 63.8 59.7 40C 0 months 84.6 100.0 85.2 3 months 72.t 8s.0 70.0 6 months 65.2 63.8 66.2 -20 0c 0 months 84.6 100.0 85.4 3 months 71.3 90.0 74.r 6 months 72.6 73.8 6s.5 -80 0c 0 months 84.6 100.0 85.4 3 months 66.6 92.5 53.8 6 months 68.5 77.5 53.6 -196 0C 0 months 84.6 100.0 8s.4 3 months 67.6 90.0 58.0 6 months 72.7 76.3 67.9 Probability Temperature <0.001 NS NS Time <0.001 <0.001 <0.001 Interaction <0.05 ns ns )Õ B. menziesü oollen viabilitv afær storage Table 3.5 Correlation coefficient between FDA staining results and in vitro germination percentage for pollen storage over six months. 2) Storage Temperature 0C) Correlation coeffrcient (r 20 0.536 4 0.902 -20 0.875 -80 0.462 -196 0.805 )'l B. menziesü pollen viability afær sûorage Figure 3.1 Temperature (A) and rainfall (B) data recorded at the nearest weather station to the Happy Valley experimental site for 1993,1994 and 1995. 58 A Mean maximum and minimum temperature at Happy Valley in 1993, 1994 and 1995 30 I Mu 1995 20 -€- Min 1995 o -_'- 0 L -.-..ç_ Mu 1993 æ ----'.ó-- Min 1993 Mu 1994 L Min 1994 L 10 q t< 0 Jan Feb lvfa¡ Apr May June July Aug Sept Oct n'ov Dec Month B Totat rainfall (mm) at Happy Valtey in 1993, 1991 and 1995 200 I 1ss4 ø less B rse3 È E 100 L F 0 lut Feb Mar Apr May June July Aug Sepr Ocl Nov Dcc Month B. menziesii pollen viability afær storage F'igure 3.2 Temperature (A) and rainfall (B) data recorded at the nearest weather station to the Nangkita experimental site for 1993, 1994 and1995. 59 A lVfean maximum and minimum temperature at Nangkita in 1993, 1994 and 1995 I a 20 - 0 Ja¡r Feb Ma¡ þr May June July Aug Sept Oct Nov D:c Month B Total rainfall (mm) at Nangkita in 1993, 1991 and 1995 300 I 1 994 Ø 1 995 m E 1 993 200 E L 100 F 0 Jan Feb Mar Apr May June July Aug Sept Oct Nov Dcc Month B. menziesü oollen viabilitv afær storage Discussion ieÇnqtta\oC 0C), Banl 0C) or liquid nitrogen (-196 0C ) retained a germinability of around 70Vo.Fot crossing te{rrnerq\or purposes, a standard fridgËoi freezer is the most practicable method to store pollen. Storage above zero degrees is not feasible, and extended pollen storage leads to a decrease in the length of the pollen tubes obtaineÅinvitro. This may influence ttre ability of a pollen to grow a tube down the style, and effect fertilisation. Theoretically, pollen in liquid nitrogen retains is viability indefinitely provided precautions :rre taken to reduce the moisture content before storage (Stanley and Linskens 1974, Sedgley 1981). The level of liquid nitrogen must be maintained, otherwise freezing, thawing and refreezing may kill the pollen. Media for invitro germrnation of pollen have been developed for a number of genera, but there were no published reports of an optimal medium for Banlcsia. Germination was not signif,rcantly affected between temperatures of 15-25 degrees forB. menziesü, perhaps reflecting cool natwal conditions in the field during the flowering soason. The species B. coccinea and B. integriþIia were also tested, and germinated well on the in vitro medium. The FDA test appeared to be less reliable than in vitro germination. In particular, the pollen stored at 20 0C fluoresced well, but had low in vitro germinability. Correlations between the media and FDA results were good in some casos, and poor in others, indicating differences in sensitivity benveen the ¡¡¡o techniques. Poor correlation of FDA and ir¿ vitro germination has been shown in Elocharis, Lonicer¿ (Shivanna and Heslop- Harrison 1981) andAcacia (Sedgley andHarbard 1993). FDA tests two propefies of the pollen grain, the integrity of the plasmalemma of the vegetative cell and the presence of esterase capable of cleaving the fluorogenic ester fluorescein diacetate (Heslop Harison et a1.,1984). Thus, FDA tends to overestimate the germinability of the pollen. It is 60 B . menziesü pollen viability afær sûorage however, considered to be more accurate than other forms of viability testing such as lactophenol, fuschin, acetocarmine and benzidine, as tested on Solanum pollen (Heslop- Harrison and Heslop-Harrison 1970). Floret position on the inflorescence had no significant effect on pollen viability, so pollen can be collected from any part of the inflorescence, as long as the pollen is fresh. The optimal time for pollen collection was mid to late season, perhaps reflecting a climatic influence of temperature as the season progresses from winter to spring. The results of this study indicate that Banlcsíapollen can be stored relatively cheaply in an fgÇt\one cc,\or ordinary household trdgÐ:or freezer for long enough to suit breeding requirements, after drying and storage over silica gel. ól withB Chapter Four Interspecific and intergeneric pollination with B anksía' . coccrneaR.Br. (Proteaceae) Abstract hand Interspecific and intergeneric pollen tube growth were investigated using confolled pollination of the commercially significant species Banksia coccínea, to species of B coccinea Banl section Oncosrylis, than to the section Banksia where it is currently placed. It is coccinea suggested that a new section Coccínea be erected. Intergeneric crosses of B. low with Dryandra species resulted in some compatibility, with one cfoss having results numbers of pollen tubes in the pollen presenter and upper style region' These genera in the indicate a close relationship between Banl t7b e B anlcs i a e, f alrúy Prote ac eae. 62 Intersoecific and interseneric oollination with B. coccin¿a Introduction The genus gen#sie t,f,, with ever 70 speeies; is a well knewn member of the family 8)x Banksias are cultivated for the amenity, cut flower and potted plant industries.4hqråe¡¡e-aa*aoti+e-bleens*and-fast"'-ryel-tr-as-ou.t-.-flowere-op Breeding is cunently underway to select new cultiva¡s for the cut flower indusqy, and interspecific hybridisation has recently been assessed as a potential breeding tool (Sedgley et a\.1991,1994,1996). Banksias prod.uce inflorescences composed of numerous hermaphrodite flowers clustered in pairs a¡ound a central woody core. Each floret has a single elongated pistil with two ovules in a unilocular ovary. Anthers a¡e attached near the tip of each of the four perianth parts by short filaments. This region of the perianth, prior to anthesis, encloses the distal portion of the style, which in many species has an obliquely terminal stigmatic groove (Sedgley et al. 1993). This portion of the style is modified for the special function of pollen prcsentation. Anthers dehisce prior to anthesis, depositing pollen onto the pollen presonter, and at anthesis the style is released from the perianth and pollen is available to foraging fauna. The stigmatic groove becomes receptive after the pollen has been removed, when it is ready to receive pollen from another plant (Fuss and Sedgley 1991). Following fertilisation, follicles and seeds may develop, taking from a few months to two years to mature. To obtain interspecific hybrids, the taxonomic relationships of the species must be known in order to increase the likelihood of a successful cross. It is assumed that closely related species will hybridise, while distantly related species will not. Taxonomic relationships within genus Banksia however, are still un¡esolved. In the current classification Banlcsia is divided into two subgenera, Banksia (72 spp.) andlsosrylis. (3 spp.) (George 1981, 1988). Subgenus Banlæía is divided into two sections, Banl flower form. Section Banl 63 and interseneric oollination with B. coccinea secrion OncosryIis contains species with hooked styles (22). V/ithin the larger secúon Banksia there a¡e ten series, and section OncosryIis has three series. Banksia and bc¿nkttery- Dryandra, are sister taxa in thembeÐadæiae,õf the family Proteaceae. Self incompatibility is common in the genus (Ramsey and Vaughton 1991), and inhibition of pollen tube growth in the pollen presenter has been shown ln B. coccinea (Fuss and Sedgley 1991), B. prionor¿s and B. menzíesii (Sedgley et aI. 1994). Mechanisms of both intraspecific and interspecific incompatibility remain obscure for Banksia, although pollen tube inhibition in the style indicates a gametophytic system (Fuss and Sedgley 1991). Further evidence for gametophytic incompatibility arises from the binucleate pollen grains seen in Banksía and other genera in the Proteaceae (Brewbaker 1959). Binucleate pollen grains commonly characterise incompatibility systems of the gametophytic type in which the inhibition occurs during pollen tube growth in the pistil. Interspecific hybridisation is of interest from economic, ecological and taxonomic viewpoints. Economically, the potential of hybrids may exceed that of the parental species for novel or improved varieties in ornamental horticulture. Hybrids occurring in the wild may have a competitive advantage over both parental species in a disturbed habitat (Potts and Reid 1985). Hybridisation generally occurs only between related tæca and can therefore have systematic and evolutionary implications for a group (Erickson er al. 1983). Some presumed natural interspecific hybrids have been reported in Banl (Taylor and Hopper 1988) but littte information is available on the limits to interspecific hybridisation. Crossing studies of Banksia have encouraged speculation on the relationships of species within the genus. Iæwis and Bell (1981) crossed fow Banksia species and found that species of the same style morphology crossed. Examples arc B. menziesü andB. attenuata in section Banlæia, which have snaight styles and^8. littoralis and.B. telmartaea in section Oncosrylis,which have hooked styles. There appeared to be a ba¡rier between the two style morphology groups, with species having straight styles & ând interseneric oollination with B. coccitua unable to cross with hooked style species. However, pollen germination was observed only in the stigma, and this is only an initial step towards obtaining a successful hybrid. In a study by Sedgley et aI. (1994), crosses were conducted involving two Western Australian species, B. prionot¿s and B.menziesÍi. It was found that success of pollen tube gtowth was largely related to taxonomic distance between the species, and the results showed a close affînity between the series Banksía and Crocin¿¿. Pollen tube growth appeared to be controlled in the pollen presenter and upper style regions. Banksia coccinea R. Br. (scarlet banksia) is currently the only species in Banksia, subgenus Banksia, section Banksia, series Coccíneae. It is placed in subgenw Banksia as its styles arc straighr at anthesis. However, prior to anthesis the styles a¡e hooked as in section Oncostylís, series Spicigerae. Placement of B. coccínea within the genus is unclear as it has no obvious close relatives having unique seedling leaves, adult leaves, inflorescences, pollen, infructescences, follicles and seeds. A recent cladistic analysis of Banlcsia based on morphological characters was undertaken by Thiele (1993); he noted however, that his analysis should be treated with caution, as support for some nodes was tenuous. In this analysis B. coccinea and other taxa were problematic and were placed incerte sedís. Additional independent data are needed to resolve these relationships. This study investigated interspecific and intergeneric compatibility in Banksia, by studying pollen tube growth using fluorescence microscopy. In particular, the study focused on the relationship of the commercially significant species Banlcsía coccínea', with othe¡ species in the genus Banksia, for the production of potential comme¡cial hybrids. ó5 ând interseneric pollination wittr B. coccinea Materials and methods Plnntmaterial Species of Banksia and Dryandra were located in collections at Happy Valley and Nangkita, South Australia. (latitude 35010' S, longitude 138o34'E). Climate and soils data are presented for the nvo sites in chapter three. All experimental plants were grown from seed, and ranged in age from 3 to 15 years. Selected species represented two genera, two subgenera, two sections, and twelve series (Table 4.4). Species were chosen on the basis of taxonomic position, flowering time, and commercial potential. Classification follows that of George (1981, 1988). Pollinations for assessment of pollen tube growth Inflorescences were prepared for experimentation (Plate a.1) by removing the open florets and isolating the inflorescences with glassine bags secured with a twist tie (Fuss and Sedgley 1991b). One day later, self pollen was removed from open florets by passing a looped synthetic pipe-cleaner over the pollen presenter region of the style, after which the inflorescence was rebagged. Approximately 100 florets were pollinated 3 days after anthesis at peak stigma receptivity (Fuss and Sedgley 1991a) with fresh pollen, on at least two inflorescences per species . B. coccined was the male parent in all interspecific and intergeneric crosses as the unique pollen grain morphology allowed easy identification of cross pollen from chance self pollination (Sedgley et al. 1993). Reciprocal crosses were selected to study possible effects of unilateral incompatibility on pollination success. Pistils were fixed in Carnoy's solution at 6 days after pollination. They were transferred to 907o ethanol for storage, before processing for squash preparations for fluorescence microscopy (Fuss and Sedgley l99la) (Plate 4.2). Counts were made of the number of pollen grains in the stigmatic groove, pollen tubes in the pollen presenter, and in the upper and lower regions of the style. Pollen tube data were 66 and interseneric pollination with B. coccinea anú\r1.,t." o\ Àcu,.u,ce \..¡eÀ ov\ 0\e4Qrc\ì.rÀ \\'ne". -or0ds analysed by ANdVA using the stâtistical packäge, Genstat 5 (Payne 1987). Experimental interspecific pollinations were conducted throughout the 1993 and 1994 seasons, and intergeneric pollinations were conducted in 1995. Results Effect onpollen tube growth of site, month andyear A subgroup of the interspecifîc crosses, tested at both Happy Valley and Nangkita, showed a signifrcant site effect (Table 4.1). Pollinations conducted at Happy Valley were more successful than those at Nangkita for all species exceptB. hookeriana. There was a significant effect of time of pollination during the season on intraspecifrc B. coccinea and B. menziesÍi crosses, but all showed successful lower style penetration (Tabte 4.2). There were differences in the extent of pollen tube penetration duringthe 1993 and L994 flowering season for the species B. ericiþlia, B. menziesíí and B. blechniþli¿, but not for B. hookeriana, B. coccinea or B. browníi (Table 4.4). Analysis of the interspecific cross data was adjusted for the site effect \^s\vr1 .¡e*e r c'\ìsecl \iu\t.,{' o*o t\e\i. Reciprocal crosses Reciprocal crosses were conducted between B. coccinea and the species B. gardneri, B. praemorsa, B. speciosa, B. ericiþlia andB. menziesií (Table 4.3). There were no significant reciprocal effects with B. gardneri and B. praemorsa. Signif,rcant reciprocal effects with B. speciosa, B. ericiþlía andB. menziesii were observed in the stigmatic groove and pollen presenter, but not the upper and lower style. o/ Intersnecific and intergeneric pollination with B. coccitwa Interspecific an¿ intergeneric pollínation with B. coccinea as nale parent Numbers of pollen grains per stigmatic groove, and pollen tubes per pistil were low (Table 4.4), but pollen tubes of B- coccinea were observed in the lower style of the species B. coccinea, B. ericiþIia, B. micrantha andB. sphaerocarpa. A Poisson model was fitted to the number of pollen grains per cross, and analysis of deviance showed a significant effect of female parent (P < 0.001). Pollen tubes of B. coccínea were observed in low numbers in the pollen presenter and upper style regions of the species Dryandra squarossa (Table 4.4). To account pollen tube numbers observed, even in intraspecifîc cross pollen tube growth were calculated based on binomial models, relating the number of pollen tubes in the pollen presenter to the number of pollen grains on the stigma, the number of pollen tubes in the upper style to the number in the pollen presenter? and the number in the lower style to the number in the upper style. Analysis of deviance showed a significant inæraction between the female parents andB. coccíneapollen in the pollen pr,gsenter (P < (l r oPo r \to tr1 0.001), upper style (P < 0.05) and the lower style (P < 0.001). The probabifidesbased on the binomial models predicted that, in addition to species in which lower style penetration was observed, success could be expected with the species B. menziesii, B. prionotes, B. ashbyí, B, attenuatø, B. Ieavigata, B. blechniþIia, B. petiolaris, B. Iemanniana, B. ticuspis, B. nutans, B. telmatiaea and B. cuneata (Table 4.5). Results from observed and predicted pollen grain and tube numbers were averaged across the taxonomic series (Tables 4.6,4.7). Pollen tubes of B. coccínea (series Coccineae) were observed in the upper style of interseries crossss to CyrtosryIís, Prostratae, Tetragorute, Spicigerae, Abietinae, and of the intersubgeneric cross to Isosrylis. Pollen tubes were present in the lower style of interseries crosses to Spicigerae andAbietinae. Predicted pt opot\r o vr1 @iÊs.indicatedthatadditional1owerstylesuccesscou1dbeexpectedwiththe series Crocírae, Cyrtostylis andTetragonae and with the subgenus Isostylis. 6E and interseneric oollination with 8. coccinca P ollen tub e ab normalitie s Some interspecific crosses produced apparently normal pollen tubes, but others showed abnormalities. Pollen grain and pollen n¡be abnormalities were obsewed on the súgmatic surface, but most pollen tube abnormalities were observed in the pollen presenter and upper style region. There were no pollen tube abnormalities observed in the lower style (Table 4.8). The type of abnormalities included directionless and branched pollen tubes, those with thickened walls, bulbous swellings, and burst tips (Plate 4.3). 69 and interseneric pollination with B. coccineø Table 4.1 Site effect on mean pollen grain and tube numbers in interspecific crosses of. Banl Pollen tubes m Species . Site Pollen Pollen Upper Lower g¡ains Presenter style style Series Bauerina¿ B.baueri Happy Valley 2.60 0.00 0.00 0-00 Nañl'icita 1.03 0.00 0.00 0.00 Series Banl Series Crocinne B. hookeríana Happy Valley 0.09 0.00 0.00 0 00 Nangkita 0.55 0.00 0.00 0 00 Series Cyrtostylis B. praemorsa Happy Valtey 0.43 0.02 0.00 0.00 Nangkita 0.r2 0.00 0.00 0.00 Series Abietínae B. telmatiaea Happy Valley 2.30 0 48 0.11 0.00 Nangkita 2.3r 0 15 0.05 0.00 Mean Happy Valley r.32 0.09 0.00 0 00 Nangkita 0.72 0.03 0.00 0 00 Probability <0.05 <0.05 ns ns 10 and interseneric oollination with B. coccinea Table 4.2 Effect of time of pollination during the flowering season on mean pollen grain and tube numbers in intraspecific crosses of Banksia coccinea and B. menzíesíi -l'rme of pollrnaÍon Pollen Pollen l,ower x 10 August t.75 0.97 0.49 0.14 27 September 3.79 0.99 0.48 0.25 Probability ns <0.05 ns <0.05 B. menziesíí x B. menziesii 16 July 0.93 0.74 0 55 0.11 10 August 0.55 0.16 0 1 0 0.02 Probability NS <0.05 ns NS 7L and interseneric nollination with 8. coccinea Table 4.3 Mean pollen grain and tube numbers in interspecific crosses of Banksia with B. coccinea as the male parent. Cross Pollen Pollen U Lower x . cocdnea B. coccinea x B. gardneri 0.02 0.00 0.00 0.00 Probability NS ns ns ns B. praemorsa x B. coccinea 0.28 0.01 0.00 0.00 B. coccinea x B. praemorsa 0.10 0.00 0.00 0.00 Probability ns ns ns ns B. speciosa x B. coccinea 0.03 0.00 0.00 0.00 B. coccinea x B. speciosa 2.2r 0.00 0.00 0.00 Probability <0.01 ns ns ns B. ericíþlia x B. coccinea 1.85 0.33 0.19 0.04 B. coccinea x B. ericiþlía 3.39 0.19 0.02 0.00 Probability <0.01 <0.01 ns ns B. menzíesíi x B. coccinea 0.45 0.00 0.00 0.00 B. coccinea x B. menzíesii 2.98 1.54 0.r7 0.00 Probability <0.01 <0.01 ns ns 7Z Intersoecific and intergeneric pollination wiü B. coccirua grain tube numbers in intraspecific,, interrpecific.,_ Table 4.4 Mean pollen and ^1n_d intergeneric crossei pollinãted with B. coccinea pollen. Taxonomy follows George (1981, 1988). Taxon Pollen Pollen UPPer Lower Year Subgenus Ban*sia . SectionBar¡,tsia Sr¡iesSalicinae B. integrifoliaL.f. 0.04 0.01 0.00 0.00 r994 B.margirafaQv. 0.45 0.03 0.00 0.00 1993 B. paludosaR.Br. 0.69 0.00 0.00 0.00 1993 Series Granles B. grandß'Y,lilld' 0J2 0.00 0.00 0.00 1994 ffiesQrcrcinac B. quercifoliak.Br. 0.08 0.00 0.00 0.00 1993 HesBatøinac B. baueriR.Br. 1.80 0.00 0.00 0.00 1993 Series Ba¡l GerusDryandra D. pracnorsa 0.81 0.00 0.00 0.00 1995 D. sessilis 0.27 0.00 0.00 0.00 1995 D. sEtarossa 0.40 0.08 0.02 0.00 1995 Probabilitv <0.05 ns ns NS 7'3 Intersoecific and intergeneric pollination witit B. coccin¿a [,, oe o*\. o*: interspecific Table 4.5 Predicted Èrebability of pollen tube growth in intraspecific, ^and intergeneric crosses pollinated with B.coccínea potlen. Taxonomy follows George (1981' 1988). Subgenus Bønlcsia SectionBaz,tsi¿ SrrjesSalicirwe B. integrifoliaL.L. 0.0400 0.2500 0.0038 0.0000 r994 B. margirataC-av. 0.4500 0.0667 0.0004 0.0000 1993 B. paludosaR.Br. 0.6970 0.0001 0.0000 0.0000 1993 Series Grandes B. grandisWilld. 0;1200 0.0001 0.0000 0.0000 1994 WesQucrcinae B. quercifoliaR.Br. 0.0864 0.0002 0.0000 0.0000 1993 HesBauerinae B. baueriR.Br. 1.6865 0.0001 0.0000 0.0000 1993 Series Bønksic B. candolleana Meissn. r.2233 0.0001 0.0000 0.0000 1993 B. meruiesiiR.Br. 0.3020 0.0264 0.9981 0.0001 1993 0.4624 0.0002 0.0000 0.0000 1994 B. ortata F. Muell. ex Meissn. 1.0515 0.0196 0.0005 0.0000 1993 B. specioseR'Br. 0.0304 0.0002 0.0000 0.0000 t994 *iesCrocirae B.burdettiiE.G. Baker 0.1915 0.0002 0.0000 0.0000 1993 B. hook¿riarn Meisnn. 0.4396 0.0002 0.0000 0.0000 1993 0.326',7 0.0001 0.0000 0.0000 r994 B. prionotesLtndley 0.7294 0.2097 0.0769 0.0002 t993 Series Cyrtosrylis B. ashbyi E.G. Baker i.6563 0.1698 0.1482 0.0002 1993 B. attenuataR.Br. 1.1800 0.0593 0.2857 0.0001 r994 B.laevigata Meisnn. 0.5700 0.3684 0.6667 0.0001 r994 B. mediaR.Br. 0.3409 0.0002 0.0000 0.0000 1993 B. praemorsa And¡ews 0.2546 0.0266 0.0004 0.0000 1993 Sr'¡jesProstratac B. blechniþliø F. Muell. 0.9038 0.0106 0.0005 0.0000 1993 09216 0.1489 0.5000 0.0001 r994 B. gardruri A.S. George 0.0962 0.0002 0.0000 0.0000 1993 B. petiolaris F. Muell. 1.565',1 0.4581 0.0'704 0.0001 r994 B. reperu Labill. 0.0374 0.0002 0.0000 0.0000 1993 MesTetragonnz B.lemanniann Meisnn. 7.597'.1 0.1029 0.2794 0.0001 1994 SenæCoccircae B. coccircaR.Br. 2.9562 0.3272 0.4906 0.423r 1993 6.1800 0.6165 o.3176 0.2810 r994 Section Oncostylis Srcnes Spicigerac B. brownií Baxter ex R. Br 0.242'1 0.0002 0.0000 0.0000 t993 r994 B. ericifoliaL.f. 0.981I 0.r4,/2 0.8000 0.0002 1993 r.882/ 0.r't7L 0.5882 0.2000 t994 B. occidentalis R. Br. 0.3301 0.0294 0.0005 0.0000 1993 B. tricuspis Meisnn. t.5775 0.t346 0.6191 0.0002 t993 þnesAbietinae B.laricirw C. Gardner 1.0316 0.0714 0.0005 0.0000 1993 B. micrantha A. S. George 2.3301 0.2417 0.4655 0.48r5 1993 B. nuta¡sR.Br. 5.1765 0.02/16 0.1538 0.0001 r994 B. pulchellaR.Br. 0.2'772 0.000r 0.0000 0.0000 1993 B. splnerocarpaR.Br. 3.5600 0.0955 0.2467 0.66't t994 B. teltnatiaea A. S. George r.9634 0.1265 0.2154 0.0002 1993 Subgenus Isostylis B. curæata A. S. George 3;1250 0.0468 0.2632 0.0002 1993 Probability <0.001 <0.001 <0.05 <0.001 GansDryandra D. pracmorsa 0.8100 0.00001 0.0000 0.0000 1995 D. sessilis 0.2700 0.00001 0.0000 0.0000 1995 D. squarossa 0.4000 0.19512 0.0000 0.0000 1995 Probability <0.05 ns ns ns 74 Intersnecific and interseneric pollination with B. coccinea Table 4.6 Mean pollen grain and tube numbers in intraspecific, interspecific and intergeneric crosses pollinated with B. coccínea pollen averaged over tÐ(onomic groups. Taxonomy follows George (1981, 1988). Taxonomic goup Pollen Pollen Upper loyer Subgenus Banksia SectionBanlcsia Series Salicinae 0.39 0.01 0 00 0.00 Series Grandes 0.72 0.00 0 00 0.00 Series Quercírae 0.08 0.00 0 00 0.00 Series Bauerinae 1.80 0.00 0 00 0.00 Series Banlcsia 0.62 0.01 0 00 0.00 Series Crocirae 0.38 0.03 0.00 0.00 Series Cyrtostylis 0.80 0.11 0.04 0.00 Series Prostratae 0.70 0.17 0.02 0.00 Series Tetragonae 7.35 0.81 0.22 0.00 Series Coccineae 4.69 2.40 0.79 0.28 Section Oncosrylís Series Spicígerae 0.85 0.12 0.07 0.01 Series Abietinne 2.45 0.24 0.08 0.03 Subgenus Isosrylis 3.77 0.72 0.04 0.00 Genus Dryardra D.praemorsa 0 81 0.00 0 00 0 00 D. s¿ssi/is 0 27 0.00 0 00 0 00 D. squarossa 0 40 0.08 0 02 0 00 IJ ând interseneric oollination with B. coccinea ptoçor\io,,rr Tabte 4.7 Predicted preÈebiti¡y of pollen tube growth in intraspecific, interspecific and intergeneric crosses pollinated with B. coccinea pollen averaged over taxonomic groups. To Taxonomic goup Pollen Pollen Upper Lower Subgenus Banksia SecttonBanksia Series Salicínae 0.3957 0.1056 0.0014 0.0000 Series Grarúes 0.7200 0.0001 0.0000 0.0000 Series Quercínne 0.0864 0.0002 0.0000 0.0000 Series Bauerirae 1.6865 0.0001 0.0000 0.0000 Series Banksia 0.6139 0.0093 0.1997 0.0000 Series Crocinae 0.4218 0.0526 0.0192 0.0001 Series Cyrtosrylis 0.8004 0.1249 0.2202 0.0001 Series Prostratae 0.7049 0.1236 0.1142 0.0000 Series Tetragonae 7.5977 0.1029 0.2794 0.0001 Series Coccíneae 4.5681 0.4719 0.4041 0.352L Section Oncosrylís Series Spicigerae 1.0028 0.0971 0.4016 0.0401 Series Abiertnne 2.3898 0.0933 0.1 803 0.1914 Subgenus Isosrylis 3.7250 0.0468 0.2632 0.0002 Genus Dryandra D.praemorsa 0.8100 0.00001 0.0000 0.0000 D. s¿ssi/is 0.2700 0.00001 0.0000 0.0000 D. squnrossa 0.4000 0.t9512 0.0000 0.0000 't6 and interseneric oollination with .8. co ccinea Table 4.8 Pollen tube abnormalities observed in interspecific crosses with B. coccíneapollen. Pollen tube ty Percentage of qpecies with abnormality in: Stigma Pollen Lower le Bulbous swellings 1 i 4 It.4 20.0 0.0 Di¡ectionless growth 1 t.4 20.0 2.9 0.0 Burst tþ 5 .7 0.0 2.9 0.0 Branched pollen tube 0.0 14.3 2.9 0.0 -t'l Interspecific and intergeneric pollination wittr B. coccinea Plate 4.1 Interspecific hybridisation technique for Banl Figure 1 Removal of all opened florets on a B. menziesíi inflorescence prior to bagging. Figure 2 Wire cage placed on inflorescence. Figure 3 Isolation of inflorescence with a glassine bag secured wittt a nvist tie. Figure 4 After one day all unopened florets are removed. Figure 5 After removing all unopened florets, a ring of florets one day old are left for pollen removal. Figure 6 Pollen removal using a synthetic pipe cleaner, the inflorescence is then re-bagged. Figure 7 Pollination of florets after 3 days, at maximum stigma receptivity, with lresh Banl 6 days florets are han¡ested for fluorescence microscopy. 78 Interspecific and intergeneric pollination with B. coccinea Plate 4.2 Scanning and fluorescence micrographs of Banksía. Figure 1 Scanning micrograph of aB. prionotes pistil showing the stigmatic groove at the tip of the pistil (arrow), the ridged pollen presenter region (arrow), and the uppor style below the constriction of the the pollen presenter (arrow). Scale bar rcpresents 500 pM. Figure 2 Scanning micrograph of a B. sceptruÌn stigmatic groove showing papilla cells (a:row) inside the groove. Scale bar represents 50 plv[ Figure 3 Fluorescence micrograph of a disected B. priorøtes pistil søined with aniline blue. Arows show the stigmatic groove at the tip of the pistil, and the transmitting tissue in ttre centre of the pistil. Scale bar represents 500 pM. Figure 4 Fluorescence micrograph of a disectedB. pulchella ovary stained with aniline blue, showing two ovules. Scale bar represents 150 pM. Figures 5, 6 Fluorescence micrograph of normal pollen tube growth of B. coccinea through the pollen presenterregion (Fig 5), and the upper style of B. coccinea (Fig 6).Scale barrepresents 150 ¡tM. 79 f Ð l¡ Inærspeciñc and interqeneric pollination with L coccinea Plate 4.3 Pollen tube abnormalities observed n Banksia coccínea interspecific crosses using fluorescence microscopy. Figure 1 Thickened pollen tube in the pollen presenter of. B.laricínø. Scale bar represents 150 pM. Figure 2 Directionless pollen tube in the pollen presenter of. B. candolleana. Scale bar represents 150 pM. Figures 3,4 Branched pollen tubes in the upper style of .8. telmatiaea (Fig 3), andB. micanthra (Fig 4). Scale bars represent 75 pM. Figures 5, 6 Pollen tubes with bulbous tips in the upper style of.B. prionotes. Scale bar Fig 5 represents 150 pM, Fig 6, 75 ¡rM. EO ar. 9 t tr Inærsneci fic and interseneric oollination with .8. co c citua Discussion There was observed or predicted sexual compatibilty between B. coccinea and species in subgenus Banksia, section Banksia series CyrtosryIis,Prostratae, andTetragonae, section Oncostylis series Spícigerae, and Abíetinae and subgenus IsosryIís. There was no interspecific pollen tube growth with species from subgenus B¿nksia section Banksia series Grandes, Quercinae or Bauerinae. This generally agrees with the hierarchical classification of George (1981, 1988). However, the particular success of interspecific pollination of B. coccinea with B. eríciþIiø (section OncosryIis, series Spicigerae), B. micrantha and^8. sphaerocarpa (section Oncostylis, series Abietinae), as measured by pollen tube growth into the lower style, suggests a closer relationship of B. coccinea to section OncosryIis than to section Banksia, where B. coccinea is currently placed. The relationship of B. coccinea to section OncosryIís has previously been suggested by George (1981) and Thiele (1993), and the results of this study provide further evidence that there is a strong affrnity to this section. These results indicate that the crurent classification of B. coccinea in sectionBanlæiamay be inappropriate, as compatibility within the section was va¡iable, with no cross observed withpollen tube growth to the base of the style. The strong affinity of B. coccineato section Oncostylis is supported by morphological evidence, as B. coccinea has hooked styles prior to anthesis, as do the members of series Spicígerae (George 1981). Despite the strong pistil-pollen compatibility, it is not appropriate to place B. coccinea in section OncosryIis as all its members have hooked styles at anthesis, whereas B- coccine¿ does not, and this is a primary cha¡acter determining the section classification (George 1981). In addition, follicles of species in this section have no lateral beak after opening. Moreover, inflorescence development in section Oncostylis is basipetal (with the exception of B . nutans) in contrast to that of B. coccinea. Placement of B. coccinea in subgenus Isosrylis is also inappropriate on morphological grounds, due to /sosry/is characte$ such as inflorescence shape which is short and conical, ovoid receptacles, 81 and interseneric pollination withB. cocciruø loosely packed perianths, straight slender pistils and obliquely ovoid follicles. In view of the distinct morphology and lack of pistil-pollen compatibility berween B. coccinea and section Banksia, it is propos Coccinea, containing only one species, B. coccinea ( This study also highlights unusual aspects of the breeding systems oporating tn Banksía. Generally low pollen tube numbers are observed in Banksia compared to other genera. This may be partly due to the small size of the stigmatic groove, which restricts pollen grain numbers (Sedg\ey et al. 1993), and the morphology of the pistil, which in B. menziesií supports the growth of no more than one or two pollen tubes to the ovary (Clifford and Sedgtey I99a). The Banksiø inflorescence characteristically shows low seed set in relation to the numbers of flowers produced (Collins a¡d Rebelo 1987, Walker and Whelan 1991). Many hypotheses have been proposed to explain this phenomenon, including breeding system controls (Lamont and Barrett 1988, Goldingay et al.l99l), pollinator behaviour (Collins and Rebelo 1987), resource control (Stock ¿r aI. 1989), seed predation (V/itkowski et al. 1.991) and spatial limitation on the infructescence (Collins and Rebelo 1987, Fuss and Sedgley 1991a). The results of this study are in agreement with intraspecifrc and interspecific pollen tube data previously obtained (Fuss and Sedgley L99ta, Sedgley et al. 1994. 1996), showing that the pollen presentet and upper style are control points in pollen tube growth. The pollen presentü has a different morphology from the rest of the style, with an abundance of transfer cells surrounding the transmitting tissue (Clifford and Sedgley 1994). This indicates a region of high flux, which may influence inhibition of pollen tubes. The upper style is the next contolling region, with lower numbers of pollen tubes inhibited than in the pollen presenter. It appears that once pollen tubes pass this region they are uninhibited to the base of the style. Seed set and seedling growth are the ultimate tests of success of interspecific crosses. E2 and intergeneric pollination with,B. co c cinea Pollen tube growth in most interspecific crosses of Banks[,a appeared to be normal, as seen in intraspecific crosses, but some combinations showed abnormalities. Pollen grain and pollen tube abnormalities were observed on the stigmatic surface, but most were observed.in the pollen presenter and upper style regions. Perhaps this is due to higher numbers of pollen tubes and more pollen tube selection in these regions. No pollen tube abnormalities were observed in the lower style. Similar observations of pollen tube abnormalities have been reported in interspecific crosses of Eucalypr¡H (Ellis et aI.l99L), Solanum (Fritz and Hanneman 1989), Lycopersicon (deNettancourt et al. L973) and Rhododendron (Williams et aI. L982). Dumas and Knox (1983) reported that pollen tube morphology and callose formation are strongly influenced by the incompatibility reaction. Pollen tubes of compatible pollinations have long thin tubes with defined callose plugs, and those that are incompatible are often thick, ending in tapered, swollen or burst tips. It is possible that the pollen tube behaviour observed on the stigma in Banksia interspecific crosses may be a more immediate and severe display of the interspecific incompatibility mechanism. The mechanism of interspecific incompatibility has long been a subject of debate. Interspecific incompatibility is defined as any post pollination process preventing the formation of hybrid zygotes through absence of pollen germination or abnormal pollen tube behaviour. This phenomenon prevents gene flow between species, in contrast to self incompatibility which restricts inbreeding and establishes upper limits to outbreeding. Interspecific incompatibitity acts as a breeding barrier between sympatric species, allowing gradual speciation within populations, but prevents foreign germplasm migration via pollen from allopatric populations and isolates invaders introduced as seed. Current scientific thought on the mechanism of interspeciflrc incompatibility is divided into two main hypotheses: the incongruity hypothesis of Hogenboom (L975,1984) and the protein - glycoprotein mediated response related to the self incompatibility (SI) mechanism (Nasrallah et aL.1985, Anderson et al.1986). The incongruity hypothesis 83 Inæ¡specifi c and intergeneric pollination with,B. co ccinea maintains that pollen tube growth ceases in the style due to the inability of the pollen tube genotype to recognise and utilise the stylar cell secretions. The glycoprotein - protein mediated rcsponse has been more widely accepted, and recent molecular data also support it (Ai et al. L990, Gray et al.1991, Jahnen et al. t989). In other genera, interspecif,rc incompatibility parallels gametophytic SI, in which a stylar glycoprotein response has been found inPrunus (Williams et aI.7982,Mau ¿f al.1982),Nicotiana (Anderson er a/. L984,1936) andPeruniø (Kamboj and Jackson 1986). Commonly, unilateral interspecific incompatibility is observed in interspecific crosses. Unilateral incompatibility prevents pollen from SC species to grow down the styles of SI species (SI x SC), whereas the reverse (SC x SÐ is allowed. Unilateral incompatibility appeared to be absent in the Banksía interspecific crosses in this study, with no significant effect of cross direction on pollen tube numbers in the lower style. The apparcnt absence of unilateral incompatibility in this study may be anributable to the fact that most Banl outcrossing and SI (Scon 1980, Ramsey and Vaughton 1991, Ca.rthew et aI. 7988, Ca¡thew 1993, Collins and Spice 1986, Goldingay et al. L99L, Fuss and Sedgley 1991a). Alternatively, the crosses may be too far distant to be successful, and cross direction may be irrelevant, with some other factor(s) determining cross success. I-ewis and Bell (19S1) crossed foar Banlcsia species and found pollen germination in species of the same style morphology: for example, B. menzíesii and B. attenuata which have straight styles, in section Banlcsia, andB. littoralis andB. telmatiaea with hooked styles, in section Oncostylis. However, there appeared to be a ba.rrier between the nvo style morphology groups, with species with straight styles unable to cross with hooked style species. The results of the present study show that this is not however, a general phenomenon..B. coccinea, which has straight styles, crosses more readily with hooked style species, than species with straight styles. It is suggested that this is because B. coccinea is more related to the hooked styled species, rather than to other straight styled species. Other factors besides interspecific incompatibility may also isolate Banksia ð4 and interseneric pollination with B. coccinea species. Factors that isolate species include geogaphical separation, and environmental and ecological and genetic factors. It should be noted, that most closses are not possible in the wild, with many species geographically isolated, with differing habitat requirements and adaptations for specific pollinator attraction. These factors in combination with genetic incompatibitity may allow some crosses to hybridise while others will not, and evidence of this is the low numbers of natural hybrids that occur. Further work is required on the mechanisms isolating species, including those that a¡e geographically isolated, and those which co-occur, yet remain sexually isolated from one another. This study demonstated the close relaúonship of Banksiato Dryandra, with some pollen tube growth in the pollen presenter and upper style regions in the cross between D. sqwürosa and B. coccinea. Baril Banl studies such as DNA sequencing are needed to address these questions. In conclusion, the species relationships hightighted by this study contribute to determining the raxonomic position of the problematic species, B. coccinea.In addition, species combinations ate suggested that may be successful in the production of novel hybrids, which may have potenúal for the cut flower and poned plant industries. E5 Banl Chapter Five Banksia sect. Coccined (4.S. George) T. Maguire et al., (Proteaceae). A new section Abstract A new section of. Banksia, soct. Coccínea (4.S. George) T. Maguire et aI., is described to include the single, but phylogenetically isolated species, Banksia coccinea R. Br. Sectional classification is based on the distinct morphological features of B. coccínea and its pollen - pistil compatibility to other species of Banlcsia. Interspecifrc pollination shows greatest compatibility with species of sect. Oncosrylis, rather than sect. Banksia, where B. coccineø is cunently placed. Due to the lack of lower style compatibility with sect. Banksia and B. coccinea's distinctive morphology, it is proposed to erect a new monotypic section as sister to sect. Oncosrylis. Introduction Tåe€cnus4 +nksiøÅs-endernie i'n-'4us*alia; -wit'L+the-exe @ extcnds--to-.P*apua--Ne'w -Guinea,-.Irian-Jaya, and_ the.,-Aru Islands,-In*the-*surre$t .elasaifieati sn; B anksic-is divided-intotwo-subg enew.-Bailæia'\1 ffi spp){Geeçge- .1981,. 1-988)"--subgenus- Bøales-iø is=further-div'iM; Banlæiaæè€neesrylis;åaseèo*flswerfem+.$eetie+9ø s*aight-styles;+r+é+eetien€*easryJt+-esnteins 22 speeies wi*l n 14+edce-witlhi¡rseet-Bs,nlcsi¿-and+hree+*scet Qneosryl.ls-. Banksia coccinea R. Br. (Scarlet barksia) is currently the only species in Banlæia subg. Bøn&siø sect. Banksia ser. Coccineae. Although it is placed in subg. Banksia, as its styles are straight at anthesis, the styles are hooked prior to anthesis as in sect. Oncosrylis ser. Spicígerae. ðó Banl The placemenr of B. coccinea within the genus has been problematic. It has no obvious close relatives and has a unique combination of characters comprising cotyledons, seedling leaves, adult leaves, inflorescences, pollen, infructescences, follicles and seeds. B. coccineø cotyledons are cuneate-obovate, with auricles (George 1981). Banksía seedling leaves are usually different from the adult leaves, and in B. coccinea the frst seedling leaves are small, broadly spathulate with na:rowly attenuate bases, deeply lobed margins, with a single obtuse lobe on each side and rounded sinuses (Thiele 1993). The broadly oblong-obcordate adult leaves are also distinctive in the combination of size, shape and marginal dentation (Thiele 1993). The inflorescences of B. coccineahave looped styles before anthesis, resulting from unequal growth of the pistil and perianth. The flowers are spirally arranged, but once the perianths are exserted they assume a vertical alignment. In most other species of subg. Banl In each pair of flowers the styles twist in opposite directions so that one overlaps with a flower of the adjacent pair. This is also the panern adopted in ser. Spicigerae. B. coccinea pollen is unique in the genus in having an elongate cylindrical shape rather than an oblong or crescent shape (Sedgley et al.1993). The infructescence of B. coccinea is relatively small and the follicles are the smallest in the genus. At dehiscence the valves of the follicte split at the insertion point of the style to leave a lateral beak. The seeds have unique features including the small size, cuneate shape, unevenly rugose inner face, unevenly wrinkled outer face and a slightly oblique, obovate, wrinkled wing. A recent cladistic analysis of. Banksia based on morphological cha¡acters was undertaken by Thiele (1993). In this analysis, B. coccinea and some other t¿xa were problematic and were placed incertae sedis. The analysis should be treated with caution, as it was noted that support for some nodes was tenuous. Because additional independent data may help öi Banl Materials and methods W Banksias-produee-inflsreseenccs-€smposeésÊ*urne*eusåerrraphrediteflewerseh*sæ+eé in-pairs around .a. cental -woody- core.- Each.-floret .has -a. +ingle olongate-pis.til+,ith-"trvo = cvdes-in+he-unilocr¡lapovary",,-A,nthors.a¡.e-attached-by-chon-f,'r.larnonts"near+trc-tipof'eaeh- of+he.{.swperianth-pa¡,çç- This r"ogion-of*the-perianth;.prior-to,-anthes'i$-eae}oses-ths [email protected],-which.-in^.rnost- sp eeies åas¿n-obli@ef y-€minaf-stigmatie grooveJhis-portion.of-the-st-ylois"modified-for',tho.f,imotion"of pollen-presentatio+P$er ffi crs-dehisee¡depositing1ol-len-ontetheaollenaresenter-. Ast¿¡fhesis +he+tyle-i+relcaeed-fromdhoperianth-and pollcn-is.a,vaitrable to .foraging-faar¡a*Bønksia €owsr+".are*protandrou+,+o.the-pollenj,s.-rnatu¡o'bofsre-theetigna.tie-groove-bceomes +eeeptiveJhe-.stigmatic*groovo-.becornes-reeeptive¿f+er-3-da"ysr-when'-itis-*eady-to ¡eeeive+ollen."from. another plant (Fuss and-sedgley- 1991),.Fotrlow'ing*fert-ilisatiorÞ, Sollieie+and+eedsnay"'dovelop;+aking"from-afew-rnoû&sto-?years-tofliattrs I nter sp e cifr c p ollínatí o n The technique for interspecific pollination was that used in manipulative intraspecific pollination (Fuss and Sedgley 1991). Inflorescences from one species of subg. Isosrylis and34 species of subg. Banksia (Table 5.1) were prepared for experimentation by removing opened florets and isolating the inflorescences inside glassine bags secured with twist ties. Between one and six species per series were selected as female parents. Self pollen was removed from open florets by passing a looped pipe cleaner over the pollen presenter region of the style, after which the inflorescence was rebagged. 88 Banl Approximately 100 florets were pollinated three days after anthesis, the time of peak stigma receptivity (Fuss and Sedgley L99L), with fresh pollen from B. coccinea, applied to at least 2 inflo¡escences per species. B. coccineawas used as the male parent in all crosses as the unique pollen grain shape ciearly distinguished interspecifrc pollination from any chance self pollination. In addition, pollen grains oî B. coccinea were able to physically fit in the stigmatic grooves of all other Banksia species (Sedgley et al. 1993). Insufficient B. coccíneø inflorescences to use as female p¿ronts for the large number of pollinations required was a further important factor. Control intraspecific pollinations were carried out on 2 species (8. coccinea and B . menziesíi) for pollen removal success' and pollination success to confirm validity of the technique- Pistils were fixed at 6 days after pollination in Carnoy's solution (6 : 3 : 1 ethanol : chloroform : acetic acid) for one week. They were transferred to 90Vo ethanol for storage before processing for fluorescence microscopy. Pistils were hydrated in 7O7o ethanol (30 mins), 30Vo e¡hanol (30 mins) followed by two washes in reverse osmosis (RO) water (30 mins each). Pistils were softened in 0.8 M NaOH at 600C (30-45 mins) and stained in decolourised aniline blue (0.17o W.S. aniline biue in 0.1 M Na2K3PO4) overnight. Aniline blue stains caliose in the pollen tube walls. Pistils were dissected longitudinally under the microscope to facilitate observation of pollen tubes, mounted in80Vo glycerol on microscope slides and covered with a coverslip. Observations were made of the number of B. coccíneapollengrains andpollen tubes in the style. Results Most interspecific crosses had apparently normal pollen tube growth down the style. Some interspecific combinations showed abnormal pollen tube growth, however, the frequency of abnormal pollen tube behaviour was low. The control intraspecific cross of B. coccinea x B. coccine¿ showed pollen tube growth to the base of the style (Table 5.1). Some interspecific combinations also had pollen tube growth to the base of the style. 89 Banksia se,cL Coccilßa Those species with lower style penetration by B. coccinea pollen tubes were from series Spicigerae andAbietinae (sect. Oncosrylis, subg. Banksia). Pollen tube growth in the upper sfyle, but not the lower style, was observed in species from the series Salícínae, Banksia, .Crocinae, Cyrtostylis, Prostratae andTetagonae of sect. Banlcsia and with subg. /sosrylís. There was no interspecific pollen tube growth with species from series Grandes, Quercinae or Baucrinae of sect. Banlcsta (Table 5.1). 90 Banl pollen- in Banlcsia pistils Table 5.1 Pollen rube pene and intraspecif,rc *-: pollinated with B. coccinea senter, us: upper style, ls: lower stylø þo[en tubes present, blank sent). Taxonomy follows George (1981' 1988). Pollen tubes inr Taxon us ls Subgenus Banksia ' SecuonBønlcsia SenesSalicinac B. integrifoliaL.f. :* B.margiranC-av. * B. paludosaR.Br. SenesGrandes B. grandisWílld. Serie.s Qucrcinae B. quercifoliaR.Br. Senæ Bauerinae B. baueriR.Br. Seies Ba¡úsia B. candolleana Meissn. B. meruiesiiR.Br. t< * B. ornataF, Muell. ex Meissn. * B. speciosaR..Br, SettesCracinac B. burdenü E.G. Baker B. hoolærianaMeisnn. B. priorøtesLndley :ß * Series Cyrtasrylts B. ashbyi E.G. Baker ¡1. * B. atterutatak.Br. * * B. Iaevigata Meisnn. t( ¡* B. mediak.Br. B. pracmorsa Andrews * Senes Prostratoc B. blechnifolí¿ F. Muell. * * B. gardncri A.S. George ¡t B. petiolaris F. Muell. * B. reperuLabill. SuiesTetragorwe B.le¡tanniaru Meisnn. * * SeiæCoccinea¿ B. coccineaR.Br. * * * Section Oncostylis Saies Spicigerae B. brownü Baxter ex R. Br. B. ericifoliaL.f. * i. B. occidentalis R. Br. * B. tricuspis Meisnn. * * SeÅæAbielinac B.IaricinaC. Gardner B. micrantln A. S. George * * {c B. nzra¡u R. Br. * * B. pulchellaR.Br. B. sphaerocarpa R. Br. * * * B. telmatinea A. S. George * * Subgenus Isostylis * B. cuncatq A.S. George 9I Banksia sæL Coccineø Discussion Interspecifrc hybridisation is of interest from economic, ecological and taxonomic viewpoints. Economically, the potential of hybrids may exceed that of the parental species for novel or improved varieties in ornamental horticulture. Ecologically, hybrids occurring in the wild may have a competitive advantage over both parental species in a disturbed habitat (Potts and Reid 1935). Hybridisation generally only occrus between related tura and can therefore have systematic and evolutionary implications for a group (Erickson et al. 1983). Some presumed natural interspecifrc hybrids have been reported in Banl between the species, and the results showed a close affinity between the series Banlcsia and Crocinae. Pollen tube growth appeared to be controlled in the pollen presenter and the upper style regions. Crossing studies with Banksiahave encouraged speculation on the relationships of species within the genus. Lewis and Bell (1981) crossed four Banksia species, observing pollen germination on the stigma, and found that species of the same style morphology crossed, vrz B. menziesii and B. attenuata which have straight styles and belong in section Banksia, and8. Iittoralis andB. telmatiaea, which have hooked sryles and belong in section OncosryIis. They suggested a ba¡rier between the two style morphology groups, with species having straight styles unable to cross with hooked style species. The current study was conducted on the basis that pollen tube compatibility tn Bartlaía is largely related to taxonomic distance, as shown previously by Sedgley et aI. (1994). This study shows that.B. coccinea has the strongest interspecific pistil-pollen compatibility with species in subg. Banlcsia sect. Oncosrylis as demonstrated by successful pollen tube growth to the base of the style. This suggests that the cunent classification of B. coccinea 92 Banksiø æcL Coccirua in sect. Banksía is inappropriate, as although compatibility within the section was variable, there were with no crosses having pollen tube gtowth to the base of the style. The strong affrnity to sect. Oncostylis is further supponed by morphological evidence, as B. coccinea has hooked styles prior to anthesis, as do the members of series Spicigerae andAbietinae (George 1981). However despite the strong pistil-pollen compatibility, it is not appropriate to place B. coccinea in sect. Oncostylís as all its members have hooked styles at anthesis, whereas B. coccine¿ does not, and this is a primary cha¡acter deærmining the sectional classification (George 1981). In addition, follicles of the species in this section have no lateral beak after opening, and inflorescence development in sect. OncosryIis is basipetal with the exception of B. nutans, in contrast to that of B. coccinea where it is acropetal. Placement within subgenus /sosrylrs is also inappropriate on morphological grounds due to characters such as inflorescence shape, which is short and conical, ovoid receptacles, loosely packed perianths, straight slender pistils and obliquely ovoid follicles. In view of the distinct morphology and lack of lower style pistil-pollen compatibility between B. coccinea and sect. Banl Coccinea containing only B. coccinea.It is further suggested that section Coccirrca is the sister sectionto Oncosrylis, based on lower style compatibility with B. cocciræa. Taxonomy Banlcsia subg. BanlcsÍa sect. Coccinea (A.S. George) T. Magutre et al,. sect., stat. nov. Banl Type specie s: B anksia coccinea R.Br. Derivation of name. From the Latin coccineus (scarlet), in reference to the styles of the type (and only) species. Stems erect. Leaves broadly oblong to obcordate. Common andfloral bracts linear- subulate. Flowers at anthesis vertically seriate. Pistil strongly curved laterally before anthesis, afterwards sfaight; pollen presenter conical. Follicles very small, after dehiscence with a lateral beak. CoryIedons cuneate-obovate. 93 Banl This section contains only one species, Banlcsia coccínea R. Br., which warrants its own section in the distinguishing characters described above. 94 Intersoecific seed set Chapter Six Seed set following interspecific pollination with B. coccineø R. Br Abstract Following analysis of pollen - pistil compatibility of B. coccíne¿ to other Banksia species, species combinations with successful pollen tube growth were selected for seed set experiments. Of the five species combinations tried, two crosses initiated follicle set. In the cross B. micrantha x B. coccínea all seed were aborted, indicating post fertilisation selection. The cross B . ericíþlia x B. coccinea produced seed. These seed showed low viability, with only two seedlings produced, one of which died. The remaining seedling, based on early seedling morphological characters, cannot be distinguished from the female parent.B. eríciþlia. Introduction Banksia breeders aim to produce novel hybrids which are faster growing, high quality, and give increased yields for the cut flower ma¡ket. Heterosis, through interspecific hybridisation, has been reported in other crops (Meskimen and Franklin 1984) but some studies show no increase in heterosis with hybrid growth rates being intermediate or equal to those of the parents (Potts et al. 1987). Moreover, some crosses give rise to inferior offspring, as seen in some eucalypts (Potts et aL 1987). Breeding of novel banksias depends on production of advanced generation hybrids, which is determined by interspecifc hybrid fertility and fitness of the subsequent offspring. 95 Interspecifrc seed set To produce interspecific hybrids in Banksía, species relationships need to be assessed. The accepted classification of. Banksia (George 1981, 1988) includes some species whose positions are uncertian. One of these is Banlcsia coccinea, a major commercial species, which has showy red blooms on long straight stems, with foliage that does not obstruct the bloom, making it ideat as a cut flower. In a previous chapter interspecific pollination was conducted with B. coccined to other Banksia species to assess sexual compatibility. Following interspecific pollination, some combinations wero found to have compatible pollen-pistil interactions, or a predicted probability of compatibility. Some self incompatibility and interspecif,rc isolation systems have been shown to operate at the post fertilisation stage (Franken et al. 1988, Seavey and Bawa 1986), and maternal resogrce allocation can also influence which seeds develop (Stephenson and Bertin 1983 Stock et al. 1989). The extenr to which interspecific isolation occurs post zygotically in Banksia can be assessed by seed set, seed germination and seedling gowth. The aim of this study is to determine the probable success of interspecific seed set following pollination with B. coccinea to species that were identified as compatible, or potentially compatible in the pistil - pollen interaction. Materials and methods Plattt møterial Following analysis of pollen-pistil interactions, a further series of interspecific túa3 pollinations wersconducted fo¡ seed set on species combinations that were identihed as compatible, or predicted to be compatible. Parental species and locations were the same as those previously described. 96 Intersoecific seed set Interspecífrc seed set Fresh B. coccineø pollen was as described previously. Two thirds of e pollinated. Female species included B. menziesii (section Banksia), B. fficuspís, B. ericiþlia, B. micrantha (section Oncostylís), andB. cuneata (subgenus IsosryIis). A controlpollination with B. coccinea was included. After pollination, bags were left in place until follicle maturity. Cones with follicles wers collected alter 12 months, at maturity. Measurements of follicle set, and length, width and height of follicles were taken. Follicles were opened by burning over a bunsen burner in a fume hood; seed and tr," **oyffiSjfu"r" removed using tweezers. Seeds were classed as full seed, or as aborted seed which were thin and papery, containing no embryo tissue. Full seeds were measured for seed weight, length, width and wing width. Differences in follicle and seed characters were tested benveen species using analysis of variance (ANOVA) with the statistical package Genstat 5. Germiration trials Seed viability was tested in sterile conditions. Seed were surface sterilised tn707o ethanol for 2 min, Milton's antibacterial solution for 5 min, followed by 3 washes in reverse osmosis (RO) water. Seeds were tansferred to sterile petri dishes with moist filter paper, sealed and kept at 15 oC in an incubator and scored weekly for germination. Seeds were scored as germinated when the radicle emerged from the seed and was about 0.5 mm long. After 3 months any ungominated seed was squashed to test for the presence of a white embryo, confirming full seed development. Differences in seed germination time and viability were tested between species using ANOVA. After germination, seeds were uansferred to potting soil (State Flora mix), placed in a glasshouse (maximum 25 oC, minimum 15 oC) and fertilised after 2 months with slow release fertiliser (Osmocote). 97 Inærspecifrc seed set Hybríd verifrcation Cotyledon morphology of newly germinated parental and hybríd seedlings were recorded. Morphotogical measurements were taken from 4 seedlings of parental and hybrid seedlings (with the exception of hybrids as there was only one surviving individual). Measurements were taken from l. leaf from the first and fifth pairs of leaves. When leaves were altemate measurements were taken from the uppermost leaf. Variables measured included: cotyledon length, width, length/width, number of leaf lobes, leaf length, width, length/width, number of teeth, length of teeth, number of nodes, distance from the first to the fifth node, hairs on upper surface, lower surface of the leaves (0 = absent, | = prosent), stem hairiness (0 = absent, 1 = present), leaf arrangement, opposite, spiral (0 = absent, 1 = present). To verify seed parentage from controlled pollinaúons data from the rwo parental species and putaúve hybrid were analysed using the statistical package PATN (Belbin 1991). Results Interspecific seed set Successful interspecific crosses were B. ericiþIia x B. coccinea and B. micrantha x B. coccinea (Table 6.1). The conüol cross also had follicte set. Interspecific pollination of B. micrantha with B. coccine¿ showed a significant difference from intraspecific pollination for follicle length (<0.001), width (<0.001), height (<0.001)' and number of seed per follicle (Table 6.2). Pollination of B. ericiþIiø with B. coccineø showed a significant difference from intraspecific pollination for follicle length (<0.001), width (<0.001), height (<0.001), seed length (<0.001), width (<0.001), and wing width (<0.001) (Table 6.3, Fig 6.1). There was no significant difference between species in number of seed per follicle, ot seed weight. 98 Interspecific seed set Seed germinatíon oÍ B. coccinea, B. ericifolia and B. coccinea -r B. ericifolia There was a significant effect (<0.001) of species combination in germination percentage, under the. conditions of the experiment, and a marginally significant effect (<0.07) of time to germination (Table 6.4). There was no significant interaction of species and germination time. S e e dli ng surtív al and v i g o ur The cross B - ericiþlía x B. coccinea gave germination and growth of two seedlings. One seedling died within 1 month of germination. Healthy cotyledons emerged from the seed, and, when planted out the seedling emerged from the soil and stopped growth. After a period of no further glowth, the growing tip senesced before production of the first leaves and the seedling died. The second seedling continued to grow and appeared to be healthy (Fig 6.2). Hybrid verification and early seedling ch¿racteristícs Cotyledon shape is characteristic of taxonomic groups within Banksía. The presumed hybrid had a cotyledon shape similar to the female parent B. eríciþIia Qig 6.3). Leaf morphology in many Banksía species changes from seedling to adult, so leaf measurements were taken at the fust and frfth leaves to standardise leaf measurements. A matrix of variables msasured for each species (Table 6.5) was used in distance calculations (Gower association measure) and subjected to group average (UPGMA) devrd¡oq¡a^ clustering using the program PATN (Belbin 1991). The dcnde$+* (Fig 6.a) shows relationships between the parental species and the presumed hybrid, which clusters with the female parenr B. ericiþIiø. The data matrix was subjected to multidimensional scaling, with a minimum spanning trse, using PATN. The data were clea¡ly separated into two dimensions (Fig 6.5) with a stress value of 0.05. A two dimensional plot of the 99 ùìì 7 c data showed spatial relationships between individuals of each species and the hybrid. The hybrid grouped with the female parent B. ericiþIia, indicating that it may be a self. Group statistics were calculated by PATN for each of the morphological characters (Table 6.6). The relative contribution of each cha¡acter was evaluated for separating the data into two groups (8. ericiþlia, B. coccinea). A probability value calculated using the Kruskal - V/allis test showed the level of significance for each character. Cha¡acters 4, 5, 7 ,8,9, 10, 11, 12, 13, 14, 15, 17, 18, and 19 were significant. r00 Intersoecific seed set Table 6.1 Number of fertile cones and mean number of follicles/cone following interspecific pollination with B. coccineapollen (SE = standald error). Interspecific cross No fertile No ba:ren Total Follicles/ cones cones cone (SE) B. menzíesü 0 4 4 0 B. tricuspts 0 6 6 0 B. ericíþIia 2 2 4 12 Q.o) B. micrantha J 1 4 16 (l.o) B. cuneata 0 8 8 0 B. coccinea 4 0 4 22 (0.6\ 101 Interspecific seed set Tabte 6.2 Mean follicle and seed set data for parental species and putative interspecifrc hybrid,B. micrantha x B. coccinea . Character B.mícrantha B.micrantln B. coccinea Probability open x x pollinated B. coccinea B- coccínea Follicle length (mm) 22.60 16.10 8.67 <0.001 Follicle width (mm) L7.73 13.13 3.83 <0.001 Follicle height (mm) 11.93 8.r7 3.40 <0.001 Seed/follicle 1.00 0.00 1.00 ns toz Inærsoecific seed set Table 6.3 Mean follicle and seed set data for parental species and putative interspecific hybrid B. ericiþlia x B. coccinea . Character B. ericfolia B. ericiþlia B. coccinea Probability open x x pollinated B. coccinea B. coccinea Follicle length (mm) 22.70 19.45 8.25 <0.001 Follicle width (mm) 10.25 9.95 3.65 <0.001 Follicle height (mm) 8.90 9.00 3.10 <0.001 Seed/follicle 1.00 1.00 1.00 ns Seed length (mm) 8.55 8.70 6.20 <0.001 Seedwidth (mm) 5.65 6.s0 4.35 <0.001 V/ing width (mm) tr.70 10.65 6.40 <0.001 Seed weight (g) 0.025 0.025 0.020 NS IU3 Interspecific seed set Table 6.4 Seed germination percentage and germination time for parental and putative hybrid seed. Cross B. ericíþlia B. ericiþlía B. coccinea Probability open X x pollinated B. cocciræa B. cocciræa Germination 7o 30 7 52 <0.001 Timeto mæc germination (weeks) J 5 5 <0.070 104 set x B. coccinea hybrid' Characters Table 6.5 Seedling morphology data for 3 month old seedlings of B. coccinea, B. ericiþtia, and putative B. ericiþlia the first leaf pair, number of leaf lobes measured are: cotyledon length (CL), cotyledon width (CW), cotyledon length/width (CLAM), characters from (lTL), number of nodes (NN)' distance from (lLLo), leaf length (lLL), leaf width (lLW), leaf length/width (|LLILSD, number of teeth (1T), length of teeth (S), upper (U), or lower (L) surface, characters from node one to node five (ND), stem hairiness (SH), leaf arrangement, oppsite (O), spiral leaf hairs on the (SLLILW'), numberof teeth (5T),length of teeth (5TL)' the fifth leaf pair, numberof leaf lobes (SLLo),leaf length (5LL),leaf width (5LUD,leaf lengtVwidth Char(j,-CWCUWIL[,olLLlLWl:|J,]LITITLNNNDSHOSULSLLoSI-L5LW5lLlL5TsTL w 40 0 27 4 6.75 8 I Eric 1 tl 8 1.4 0 18 3 6 8 17 01001 40 0 '26 3 8.'l 6 I Enc2 12 7 t.7 0 20 2 10 6 18 01001 40 0 24 4 6 5 I Eric 3 9 8 1.1 0 19 3 6.3 8 l7 01001 0 24 3 8 I I Edc4 ll 7 1.6 0 14 3 4.7 8 18 40 01001 0 0 30 1 0 0 8 31 1l 2.8 Coccl 9 8 1.1 3 14 l0 1.4 0 05 0 0 1 0 0 13 38 t4 2.7 Cocc2 9 'l 1.3 3 93 3 0 05 20 t 20 I 0 0 l3 35 14 2.5 0 0 Cocc3 8 6 1.3 3 105 0 05 t7 I 0 8 19 10 1.9 0 0 Cocc4 10 9 1.1 3 93 3 0 05 101 30 0 l0 0 24 3 8 6 I E c10 9 1.1 0 192 9.5 6 r9 Interspecific seed set Table 6.6 Morphological character evaluation. Group statistics given two groups, B. ericiþIía (1) and B. coccinea (2). Character mean, standa¡d deviation and probability using Ikuskal - Watlis test, calculated by the program PATN (Belbin 1991). t. v 2 9.00 0.7t 0.0595 2 1 7.80 0.74 2 7.50 L.t2 0.7024 J 1 1.38 0.24 ,, t.20 0.10 0.3039 4 1 0.00 0.00 2 3.00 0.00 0.0047 5 1 18.00 2.09 2 10.50 2.06 0.0184 6 1 2.60 0.49 2 5.25 2.86 0.0591 7 1 7.30 2.08 2 2.35 0 68 0.0139 8 1 7.20 0 97 2 0.00 0.00 0.0088 9 1 1.00 0.00 2 0.00 0 00 0.0047 10 1 7.80 0.75 2 5.00 0.00 0.0098 11 1 38.00 4.00 2 2t.75 4.92 0.0142 t2 1 0.00 0.00 2 1.00 0.00 0.0047 13 1 1.00 0.00 2 0.00 0.00 0.0047 t4 1 0.00 0.00 2 1.00 0.00 0.0047 15 1 0.00 0.00 2 10.50 2.50 0.0067 16 I 25.00 t.27 2 30.7s 7.22 0.2129 t7 1 3.40 0.49 2 t2.25 r.79 0.0120 18 I 7.49 0.97 2 2.48 0.35 0.0139 T9 1 6.60 1.20 2 0.00 0.00 0.0098 IUO Inærsoecific seed set Figure 6.1 Schematic drawing of parental and putative interspecific hybrid seed- Seed are drawn to actual size. (a) B. eríciþlia, (b) B. eríciþlia x B. coccinea, (c) B. coccinea, (d) B. micrantha, (e) B. miqantha x B. cocctnea, (Ð B. coccinea. ro7 J d p c q e Interspecifrc seed set Figure 6.2 Putative interspecific hybrid seedling between B. ericiþlía and B . coccinea at 3 months. IOE DtilÞru|tl \ ÙltÚtru¿ 'tl Intersoecific seed set Figure 6.3 Schematic drawing of parental and putative interspecific hybrid cotyledons. Cotyledons are drawn to actual size. (a) B. ericiþ\,d, (b) B. ericiþlíø x B. coccinea, (c) B. coccinea. 109 J q e Inærspecifrc seed set Figure 6.4 Dendogram showing relationships between B. coccinea, B. eríciþlía arñ putarive interspecific hybrid. Distance matrix calculated using the Gower association matrix, followed by group aveÉge (UPGMA) clustering. 110 0.0646 0.1965 o -3284 0 .4602 o -592t 0 -7240 lt I I I B. ericifol-ia 1 B. ericifolia 3 I B. ericifol-ia 2 I B. ericifolia 4 I I _T Putative hybrid B. coccinea 1 B. coccinea 4 B. coccinea 2 B. coccinea T I I I 0.0646 0.1965 0.3284 0 .4602 0.5921 0 .7 240 Interspecifrc seed set Figure 6.5 Multidimensional scaling of data matrix into two dimensions (stress value 0.05). The plot shows spatial relationships between individulas of B. ericiþlía, B. coccittea and the putative interspecific hybrid- 111 1 2 4 r B. ericifulia Hybrid 3 cì 0 9 q') B. cocctnea 2 5 4 -1 -1 0 Vector I Intersoecific seed set Discussion The species combinations tested in this study resulted in only one cross (8. ericiþlia x B ' coccinea) with full seed formation. Of these seed only two seedlings were produced, one of which died possibly due to hybrid breakdown, and one survived. The surviving plant, based on early seedling mo¡phological characters, appears to be a self of B. ericiþlia. This does not necessarily mean it is a self, as hybrids may display dominant and recessive characters of each parent, but based on early seedling characters it cannot be distinguished from the female pÍìrent. As the plant reaches maturity, it is possible that it may display more intermediate characters. The cross B . micrantha x B. coccinea initiated folticle set, but all seed were aborted showing post fertilisation selection. Seed formation with B. ericiþlia, and partial success withB. micrantha, both in section Oncostylis, demonstrates the close compatibility of B. coccinea to section OncosryIis, as indicated previously using pollen - pistil data. Therefore, pollen - pistil interactions can predict possible hybrid combinations and species relationships. Interspecific hybridisation has previously been conducted with Banksia species to investigate species relationships Lewis and Bell (1981).erossedJeur tsadæi*speeies;-sbser+ed-pollen-tubegro$Éth-on-the-stigmaand4ounèthat*peeieroÊthe- *ame*tyl,e-morphology-e*essøüB.nenzeßüwt&B=a+teftuetaryv,*ttah-heveseaigh+styles- js-seêtioaj.44ksiøerosseü.-a&d-8"-Ii,ttora,Iisa)d,-B-+elwtlaea;+'hi@ in*seetion* Oneostylis..erossod.JFhe-re-appe{red ts-be-a-bar¡ier' between-tllo"{rvo+'tyle -morphology*groups,-with- speeiesåaving+uaight-style*'trnable teer-oss-'with-hooke+sryte- speci,es-..Elo^wever,,.this-doos not"appear.-to be-.the-"oase-in-gener+lJhe-euren+stud¡ shs¡vs-3.-eoeeinea-{suaig}rt-s.tyle) elosses -wrtí-B--erieiþJlø-and-pa*ially-with-e. ntiernfttfu(hookedstyle). Another study by Sedgley et aL (1994) involving B. prionotes and B. menziesü found that success of pollen tube growth in the style was largely related to tÐ(onomic distance between the species, and the results showed a close affinity tt2 Intersoecific seed set between the series Banksia and Crocinø¿. Thus, interspecifi.c hybridisation is a suitable technique to determine compatibility relationships benveen Banksia species. An important consideration in hybrid production for plant breeding is the Presence of pre and post fertilisation barriers. This study demonstrates that both pre and post fertilisation ba:riers occur inBanksía. Thecrosses with B. menzíesii,B. fficuspis, and B. cuneata (predicted by lower style pollen tube penetration) failed to produce follicles, suggesting selection at the ovary. The cross B. micrantha x B. coccineø set follicles, but all seed were aborted after follicle initiation. In Banl few (Taylor and Hopper 1988). Reported hybrids are usually between taxonomically close species, appear to be intermediate between the parental species, and occur in the same area. Thus, hybrid formation may be benefrcial in extending species range, species adaptation, or increasing pollinator attraction. The low occurrence of natural hybrids supports the hypothesis of selection of the fittest individuals and maintenance of genetic purity of the species. In this study the cross B. ericiþIiaxB. coccinea set both follicles and seed, however seed viability was low with only one seedling surviving which appeared to be a self. Embryo rescue has been used in other crops to recover potential hybrid zygotes before abortion and subsequent breakdown. It is possible that many of the hybrid seed formed in the B. ericiþIía x B. coccineacross resulted in embryo breakdown and were therefore inviable. In Banlcsia, little information is available on interspecific hybridisaúon and the mechanisms controlling pre and post fertilisation selection. I13 DNA isolation from B anksia Chapter Seven DNA isolation methods for Banksiø and other members of the Proteaceae Abstract The application of curent nucleic acid technologies to crop improvement require the deveþment of efficient DNA extraction techniques which yield DNA of adequate purity. Three different DNA extraction methods are described, suitable for genera in the family Proteaceae, which are both rapid and efficient. Due to the different types of plant tissue available, methods were developed for mature leaves, seedling leaves and resting buds, and seed material. These procedures combine and modify previously published techniques. The DNA is suitable forrestriction digestion andPCR amplification. Introduction The applications of current nucleic acid technologies to crop improvement include gene mapping, genetic frngerprinting, population studies, and phylogenetic analyses. These techniques are appropriate for research into banksias, proteas and other plants for cut flower production. The development of eff,rcient DNA extraction techniques, which yield DNA of a purity adequate for restriction enzyme digest and PCR, has proven difficult from plant materials in the Proteaceae. The problems associated with DNA extracúon from plants high in polysaccharides and phenolics, such as occurs in Banl recognised previously (Couch and Fritz 1990, John 1992, Do and Adams 1991). Plants contain three types of DNA, in the nucleus (genomic), mitochondria (mDNA) and chloroplasts (cpDNA). Methods of extraction for each DNA type arc available for many It4 DNA isolation from Banksia plant genera, although most experiments require preparation of only genomic DNA. Extraction procedures include removal of the cell wall and nuclear membrane; separation of DNA from the other cellular components, such as cell walls, proteins, lipids or RNA; with maintenance of DNA integrity, and protection from nucleases and mechanical shearing. Commonly, cells a¡e broken by grinding in liquid nitrogen, when the low temperatue prevents degradation by nucleases. The extraction buffer usually contains compounds such as sarkosyl, to dissolve membranes and denature proteins, and EDTA which complexes Mg 2+ ions, an essential co-factor for nucleases. After phenol/chloroform extraction the aqueous phase contains the DNA and RNA, while polysaccharides and some proteins and lipids are in the organic phase. Cell debris and protein aggregates occur in the interphase, between the organic and aqueous phases- Iso- propanol precipitation of the aqueous phase concentrates the DNA, and after a series of washes the DNA is ready to use. An optimal DNA extraction method needs to be quick and uncomplicated, allowing maximum number of extractions per day, and minimising time and cost. Developing DNA extraction methods that a¡e applicable to a broad range of species is an important goal. The aim of this resea¡ch is to develop efficient procedures for the extraction of DNA from Banlcsia and other genera in the family Proteaceae. Materials and methods Plant material from Banksia, Dryandra, Leucandendron,Isopogon, Protea, Macadamia, Leucospermum, Grevillea, Hakea and Serruria (Proteaceae) was obtained from mature leaves, seedling leaves, buds and seeds. Material can be stored for a few months at -20 0C and -80 0C indefinitely. Two methods for mature and seedling leaf, and bud material adapted from Doyle and Doyle (1983) using CTAB (Maguire et aI. 1994) were developed, and a third "miniprep" extraction method applied to seed material, is a slightly modified method of Weining and Langridge (1991). DNA extracted using the seedling 115 DNA isolation from Banksia leaves method was restricted using Hae III or Dra I endonucleases, or subjected to random amplified polymorphic DNA (RAPD) techniques (Hu and Quiros 1991, Klein- Lankhorst et al. l99L). Results Protocol 1 : DN A extaction from mnture leaves 1 Extraction buffer (5 to 7.5 rnl-) l27o (Vv) CTAB, 1.4 M NaCl, 0.2Vo (v/v) 2- mercaptoethanol, 20 mM EDTA, 100 mM Tris-HCl, (pH 8.0)l was pre-heated to 6@C in a water bath in a 10 ml plastic centrifuge tube. 2 Fresh tissue [0.5 to 1 g] was ground to a powder in liquid nitrogen in a pre- chilled mortar and pestle. 3 The powder was scraped directly into pre-heated buffer in a pre-warmed mortar and mixed thoroughly with a spatula. The sample was por:red into a centrifuge tube and inverted gently several times. 4 The sample was incubated at 6@C for 30 min with occasional gentle inversions. 5 The solution was extracted once with chloroform-isoamyl alcohol (24:1, v/v), by gentle but thorough mixing for 5 min. 6 The phases were separated by centrifugation in a swinging bucket rotor-type centrifuge at 4000 rpm at room temperature for 3 min. I1ó DNA isolation from B anksia 7 The aqueous layer was removed with a wide-bore pipette and transferred to a clean plastic centrifuge tube. Two to 3 volumes of ice-cold isopropanol was added. 8 The nucleic acids were recovered by centrifugation at 4000 rpm for 3 to 5 min. As much supernatant as possible was removed without disturbing the precipitate. TOVo ethanol was added directly to the pellet and swirled gently. If ttre precipitate was not obvious the solution was recentrifuged (this may cause some difficulty in washing and requires lifting the pellet from the bottom with a glass rod to allow thorough washing). 9 The pellet was briefly vortexed and allowed to stand for2O min in 707o erhanol and then centrifuged for 15 min at 4000 rpm. 10 The supematant was carefully removed and the pellet allowed to air dry at room temperature for approx 5 to 10 min or until there were no visible drops of liquid in the tube. 11 The pellet was resuspended in 400 pt TE buffer [10 mM Tris-HCl OH 7.4), 1 mMEDTAI. For more difficult plant material, such as bud material, this procedure was modified by incubating for 45 mins in CTAB and allowing the DNA to precipitate in isopropanol at -20oC for approx 30 mins. TL'I DNA isolation from B anksia Protocol2: DNA extractionfrom seedling leaves and other material high in phenolics and polysacclnrides (v/v) 1 Extraction buffer (5 to 7.5 rnr-) [3vo (w/v) CTAB, 1'4 M NaCl, o'47o 2- mercaptoethanol, 20 mM EDTA, 100 mM Tris-HCl, (pH 8.0)' 2VoPYPf was pre-heated to 60oC in a water bath in a 10 ml plastic centrifuge tube. 2 Fresh leaf tissue I0.5 gl was gtound to a powder in liquid nitrogen in a pre- chilled mortar and pestle. 3 The powder was scraped directly into pre-heated buffer in a pre-warmed monar and mixed thoroughly with a spatula. The sample was poured into a centrifuge tube and inverted gently several times. 4 The sampie was incubated at 6@C for 45 min with occasional gentle inversions. 5 The solution was extracted once with chloroform-isoamyl alcohol (24:1, v/v), by gentle but thorough mixing for 5 min. 6 The phases were separated by centifugation in a swinging bucket fotor-type centrifuge at 4000 rpm for 3 min at room temperatue. 7 The aqueous layer was removed with a wide-bore pipette and transferred to a clean plastic centrifuge tube. The chloroform extraction was repeated. 8 Two to 3 volumes of ice-cold isopropanol plus 50 pL sodium acetate (pH 4.8) was added to the resulting supernatant and left at -200C for approximately 30 mins. ltö DNA isolation from Banksia 9 The nucleic acids were recovered by centrifugation at 4000 rpm for 5 min. The viscous layer and pellet were precipitated v'irrth957o ethanol and centrifuged for 10 mins at 4000 rpm. 10 The pellet was dissolved in 400 Ff TE buffer [10 mM Tris-HCl (pH7.4),1 mM EDTAI with a Pasteur pipette. 1.6 mL o12.5 M NaCl was added to give a total concentration of 2 M NaCl, and the DNA was precipitated with 957o ethanol and centrifuged for l0 min at 4000 rpm. At this concentration of NaCl, contaminating material remained in solution. 11 As much supernatant as possible was removed without disturbing thepellet-70Vo ethanol was added directly to the pellet and briefly vortexed. The tube was centrifuged for 3 min at 4000 rpm. This wash procedure was repeated. L2 The supernatant was carefully poured off and the pellet allowed to air dry at room temperature for approx 5 to 10 min or until there were no visible drops or liquid in the tube. 13 The pellet was rosuspended in 100 pl TE buffer [10 mM Tris-HCl (pH 7.4), 1 mMEDTAI. Using these techniques DNA yields of 8-12 pg/g fresh wt. leaf tissue were obtained. I 119 DNA isolation from B anksia Figure 7.1 Agarose ge| (L.6Vo) showing DNA yield of Banksia cuneata individuals following DNA extraction and RNAase digestion. Standard samples of salmon sperm genomic DNA used to quantify yield. Lane 1. DNA size ma¡ker ÀHindltr (bp), lane 2. 100 ng/FI, lane 3. 75 ngl¡tL,lane 4. 50 nglpl, lane 5. 25 ng/¡tL,lane 6. 10 ng/¡tl, lane 7 . B . cuneata, Iane 8. B . cuneata, La¡re 9. B. cuneata, lane 10. B. cuneata. 120 12345678910 23 64 --- 958 6'7 42 446't l: aaoa -l-- 791 580 DNA isolation fr om B anksia Protocol 3: Isolation of total genomic DNA from seed material (slíghtly modífted " miniprep " metho d of Weining and l-angríd g e, I 99 I ) 1 One seed is frozen in liquid nitrogen and crushed in a 1.5 mL eppendorf tube with a steel rod 2 600 pL of DNA extraction buffer [47o sarkosyl, 100 mM tris HCl, 100 mM NaCl, 10 mM EDTAI is quickly added and mixed well 3 600 FL of phenol/chloroform 1257o phenol,247o chloroform, I Vo iso-atnyl- a-lcohotl is then added and the tubes are mixed by thorough hand mixing. The samples are placed on ice until all samples are ready. They are continually mixed and allowed to sit for 5 mins. 4 The samples are centrifuged at 8-10 000 rpm for 2-3 mins, then the upper aqueous phase is transferred to a clean tube using a pipette 5 The phenoVchloroform extraction is repeated 6 75 pL of 3 M sodium acetate (pH 4.8) is added with 600 pL of cold isopropanol and mixed by inversion. The DNA is allowed to precipitate for 1 mm 7 The sample is centrifuged at 12 000 rpm for 10 min and the supernatant is carefully poured off 8 1 mL of 70Vo ethanolis added to the pellet and gently vortexed TZI DNA isolation from Banksia 9 This is centrifuged at 12 000 rpm for 2-3 mins and the supernatant is poured off and the procedure repeated 10 The pellet is air dried by draining onto tissue, but not allowed to dry out 11 The pellet is resuspended in 50 pL of TE buffer [10 mM tris HCI pH 8, 1 mM EDTA pH 8l and 1 ¡rL R40 t40 glmL RNAase A in TEl, with a gentle vortex and then stored at4}Cfor up to one month or -20 0C indefinitely. GeI electroplnresis 5 pL of the DNA standards (100, 75, 50,25, 10 ng of genomic salmon spelm DNA) were mixed with 2 ¡t"L of ficoll dye and DNA samples were electrophoresed on L.5Vo agarose gels in 1X TBE buffer [5X TBE: 45 mM Tris base, 45 mM boric acid, 1 mM EDTA, ph 8.01 at2Y/cm for approximately 1.5 hours. Gels were stained with 0.5 PglmI ethidium bromide and photographed (Potaroid 667 frlm) under UV light to check for purity of the preparation after RNAase treatment, and to quantify DNA yield (Figure 7.1). Discussion The best results of DNA extraction from Banksia were with seeds and fresh leaf tissue ha¡vested from mature plants. Seedling material was more difficult as the leaves were soft and hairy, hence the development of protocol 2 suitable for seedling leaves. The DNA samples were completely digested with Hae Itr and Dra 1 restriction enzymes indicating low levels of DNA methylation (Nelson and McCleltand 1991).The DNA has also been used successfully for the random amplifiedpolymorphic DNA (RAPD) technique. rzz Genetic divers iw n B anksia and Drv andra Chapter Eight Genetic diversity of Banksia and, Dryandra (Proteaceae) using RAPD markers Abstract Random amplified polymorphic DNA (RAPD) markers were investigated as a tool for estimating genetic diversity within 33 species of. Banlcsia and three of Dryandra. Three primers were used on DNA from ten seeds per species, and band data were pooled to give between 52 and 89 bands per species, most of which were polymorphic- Genetic diversity was calculated using six published metrics on threo species, for which allozyme data were also available. Based on between method consistency, three matrices were chosen for analysis of the full data set. Levels of genetic diversity in Banksia and Dryandraranged from 0.59 to 0.90. Introduction Scientific approaches to the conservation and exploitation of plant genetic resources require a detailed knowledge of the amount and distribution of genetic diversity within a species. Traditionally, a combination of morphological and cultural traits was used, but these cha¡acters are influenced by the environment and may not reflect the true genetic diversity. To overcome these problems biochemical methods were inuoduced, and allozymes are now used extensively to characterise plant genetic resources. Allozymes have been used in Banksia to estimate genetic diversity and high levels have been reported, amongst the highest recorded for plants (Scott 1980, Schemske and Lande 1985, Carthew et al. L988, Coates and Sokolowski 1992). Large variability is expected in Bønlcsia as the mating system is predominantþ outcrossing, with pollination by birds IZ3 Genetic divenity n Banlcsia and Dryandra and small mammals. There are however, widely recognised limitations of allozymes. Many species have low levels of detected allozyme diversity, because the detection of variability is limited to protein coding loci, which may underestimate levels of genetic diversity. (Clegg 1989), and may not represent the entire genome (Schaal et al. l99L). Allozymes are also tissue specific and are influenced by environmental factors. Given these limitations many workers have moved to DNA markers such as random amplified polymorphic DNA (RAPD), based on PCR technology (V/illiams et aI.1990a). RAPDs overcome many of the limitations of allozymes and are simpler and easier to use than other DNA markers such as restriction fragment length polymorphisms (RFLPs). RAPDs have potentially unlimited numbers of markers and have been shown to be useful for population genetic studies on a number of genera (Huff et al. 1993, Chalmers et al. L992, Russell et aI. 1993). It has also been suggested that RAPDs may be an appropriate technique to monitor diversity in plant populations (Anderson and Fairbanks 1990, Waugh and Powell 1992). By using single random primers this techniquo can amplify regions within the genome without any prior sequence information, using small amounts of tissue. It is also reasonably simple and can provide a large number of potential polymorphic loci, ideal in the investigation of ra¡e and endangered plants, where there is often little material and low allozyme variability. Banl currently exploited for the cut flower industry. Many blooms a¡e harvested from the wild, although Banlcsia arrd Dryandrø species are increasingly being brought into cultivation. Some species are exploited for their foliage as well as flowers. Currently, selection and breeding for superior genotypes of. Banksia is underway to produce cultiva¡s for the cut flower industry (Sedgley et a1.,1994,1996). To assist breeding knowledge of the level of genetic diversity within a species is essential for selection and conservation of genetic resources. r24 Genetic diversiw n Banksia andDrland¡a Banl Conservation Act 1950-L979, seven banksias a¡e declared rare flora. They are B. browníi, B. chamaephyton, B. cuneata, B. goodii, B. meisnerí, B. splaerocarpa andB. tricuspís.Using the criteria for designation of rare flora in'Westem Ausnalia, another æn species of. Banksía should be added to the list (George 1987). The importance of documenting the level and distribution of genetic diversity in species in order to design optimal conservation strategies is widely recognised (Hamrick 1983, Moran and Hopper 1987). The aim of this study is to investigate levels of genetic diversity within species of Banlcsia and Dryandra ustngRAPDs, and to evaluate six methods of data analysis. Materials and methods Plarumaterial Seeds of 33 species of. Banl were obtained from Nindethana Seed Service. The species chosen represented the two sister genera Banlæia and Dryandra, and two subgenera, two sections and thirteen series within the genus Banksia (George 1981, 1988) (Table 8.1). Ten seeds of each species were randomly selected for RAPD analysis. DNA extractíon DNA was extracted from each seed using a modified method of Weining and Langridge (1991), comprising: phenoVchloroform/isopropanol extraction for 5 min on ice, DNA precipitation for 1 min with ice cold isopropanol and sodium acetate, with DNA recovered by centrifugation at 12 000 rpm for 10 min. The pellet was washed twice wittt 70 Vo ethanol, dried and dissolved in 50 pL of TE buffer, with 1.0 pL of RNAase (R40: LzJ Genetic diversity n Banksia md Dryandra 40 glmLRNAase A in TE), and stored at40C short tenn (up to I month) or -20 0C tong term (up to 1 year). DNA was subjected to gel electrophoresis on a 1.67o agarose gel in TBE buffer (Sambrook et aI.1989), and stained with ethidium bromide. DNA concentration was estimated by visual assessment of band intensitios, compared to salmon sperm genomic DNA standards. The DNA content was adjusted to 10 ng pL -1. DN A amp Iifi c atío n and do cume nt atio n DNA amplification was performed in a MJ Research Thermal Cycler. The optimal progam for Banksid commenced with an initial denaturation step at940C for 5 min, followed by 40 cycles of 94 0C for 1 min, 36 0C for 1 min, 720C for 2 mins, terminated with a final extension step aú20C for 5 min. Optimisedreaction conditions were carried outin a25 ¡tL total volume containinglXTaq buffer (Gibco-BRL), 3 mM MgCl2, 200 pM of each dNTP (dGTP, dATP, dCTP, dTAP), 1 unit of Taq polymerase (Gibco- BRL), 0.5 ¡rL T4 gene 32 protein (Boehringher Mannheim), 1 pM 10 mer primer (Operon Technologies Inc.) and 10 ng pL -1 template DNA. Each reaction mix was overlaid with PCR grade paraffin oil. DNA amplification fragments were separated by 2Vo aflarcse gel (Seakem, Promega) electrophoresis using TBE buffer (Sambrook ¿r ø/. 1989). A DNA size marker was used (pGEM, Promega) on each gel, and gels were stained with ethidium bromide. Fragment patterns were photographed under UV light with Polaroid 667 film for further analysis. Polaroid photographs were scanned using a transmission scanner (Hewlett Packard Scanjet IIc>r/T). The intensity and molecula¡ weight of each visible band was determined using the softwa¡e CREAM ttr,t., Kem-En- Tec Software Systems, Blue Sky Scientific. Sixty primers were evaluated for their suitability in a pilot survey (series OPA, OPB and OPC, Operon Technologies Inc.). Three primers were selected, these were OPA-20 tz6 Genetic divenity n Banksia and. Dryandrq (GTTGCGATCC), OPB-03 (CATCCCCTG) and OPB-04 (GGACTGGAGT) which gave reproducible and informative markers. Band fragments were scored as present (1) or absent (0) (Fig 8.1). A negative control was added in each run to test for contamination. In order to test reproducibility, the selected primers were tested three times on the same sample, for a random subset of three DNA samples. To aid interpretation of band homology between gels, each gel contained a standard DNA lane and pGEM DNA ma¡ker. The presence or absence of bands was determined for all individuals and a matrix of RAPD phenotypes was assembled. Each individual was represented by a vector of ls and 0s for the presence or absence of any particular band for all the primers used in the study. Dataarøþsis From the matrix of RAPD phenotypes for each DNA sample, an index of genetic distance (D) was calcuiated (D = 1 - F), where F is similarity calculated using six methods. Nei and Li (1979) matching coefficient method: 2x nl I / ((2*n 1 1 ) +n0 1 +n 1 0) where n=number of band positions, n11= the number of positions where x=1, y=1, n00= the number of positions where x=0, y=0, n01= the number of positions where x=0 and Y=1, n10= the number of positions where x=l and y=0' x,y = individuals being compared. The Jaccard coefficient (Jaccard 1901): nl1/(n-n00). The method of Russell and Rao (1940): n1l/n Simple matching coefficient (Apostol et al.1993): (nl1+n00)/n The method of Excoffier et aI., (1992): n*(1-(n11/n)) The method of Rodgers and Tanimoto (1960): (n 1 1 +n00)/(n1 1 +2*(n 10+n0 1 )+n00). L:N Genetic diversity tn B anlcsia and Dryønd¡a Distance matrices were calculated using the statistical package, RAPDistance (Amstrong et al. 1994). The mean distance value for each species was taken as an estimate of geneúc diversity, and the standard deviation of the mean was calculated for each data set. Results Band data for the three selected primers were pooled such that each species had total band numbers ranging from 48 to 89. The number of monomorphic bands was low, with most bands being polymorphic (Table 8.1). Preliminary analysis of RAPD data for three Banksía species with each of the six published statistical methods gave a wide range of estimates of genetic diversity (Table 8.2). The species were chosen on the basis that there were reporred estimates of genetic diversity based on allozyme elecnophoresis data (Scott 1980, Coates and Sokolowski L992). Based on between method consistency, the methods of Nei and Li (1979), Jaccard (1901) and Russell and Rao (1940) were used for analysis of the total data set. The estimates of genetic diversity for all species with the three chosen matrices were high, ranging from 0.59 for Banksia praemorsa, B. menzíesíí and Dryandra formosa, vp to0.90 forB. lemanniana andB. cuneata (Table 8.1). 128 Genetic diversitv n B anl,s i¿ nd D rl andra Table 8.1 Band data and estimates of diversity for 33 species of Banksia and three of Dryandra using RAPDs (standa¡d deviation). NUMDET OI D¿ìNOS Metnc Species Poly- Mono Toøl Nei andli Jacca¡d Russell and morphic mo¡phic Rao 190 Subgenus Banksia SectionB¿z,tsr¿ SenesSalbinac B. irxegrifolia* 82 0 82 0.77 (0.07) 0.74 (0.09) 0.86 (0.04) B. robur 61 0 6t 0.7e (0.06) 0.77 (0.07) 0.88 (0.04) Sene.sGrond¿s B. grandis 63 0 63 0.72 (0.06) 0.68 (0.07) 0.82 (0.03) B.solandri ** 52 54 0.67 (0.07) 0.62 (0.09) 0.75 (0.05) Sr,ri:esQuercinae B. Eurcifolia** 51 I 52 0.73 (0.06) 0.69 (0.07) 0.83 (0.04) SenæBeßrirce B.baueri 86 2 88 0.72 (0.06) 0.68 (0.08) 0.81 (0.05) Serie,s B¿l¡kJr¿ B. baxteri 84 0 84 0.71 (0.06) 0.67 (0.07) 0.82 (0.03) B. candolleana 60 0 60 0.74 (0.06) 0.71 (0.08) 0.83 (0.04) B. meruiesii 79 5 84 0.65 (0.06) 0.5e (0.07) 0.77 (0.04) B.serrata 53 I 54 0.73 (0.07) 0.70 (0.08) 0.84 (0.04) SeriæCræinac B. burdettii** 89 0 89 0.71 (0.08) 0.67 (0.07) 0.83 (0.04) B. prionotes 57 I 58 0.73 (0.08) 0.69 (0.0e) 0.82 (0.06) Series Cyfosfylrs B. aslþi 69 1 't0 0.67 (0.07) 0.61 (0.09) 0.77 (0.04) B. alteruata 60 I 6t 0.77 (0.07) 0.74 (0.08) 0.86 (0.05) B,audax 69 0 69 0.79 (0.06) 0.77 (0.07) 0.88 (0.04) B. elderiana 52 I 53 0.73 (0.07) 0.69 (0.09) 0.83 (0.06) B.lacvigøa * 58 0 58 0.75 (0.07) 0.72 (0.0e) 0.85 (0.05) B.Iind,eþarw * 69 2 7l 0.72 (0.05) 0.69 (0.06) 0.82 (0.03) B. praemorsa** 57 4 61 0.65 (0.10) 0.se (0.13) 0.79 (0.0s) HæProstratac B. blechnifolía ** 55 4 59 0.69 (0.07) 0.65 (0.08) 0.80 (0.04) B. repens 77 1 78 0.73 (0.07) 0.6e (0.09) 0.83 (0.04) SenesTetrøgornc B. caleyi 57 1 58 0.74 (0.06) 0.70 (0.07) 0.82 (0.04) B.letrunniana 82 0 82 0.81 (0.07) 0.79 (0.08) 0.90 (0.04) SenesCocciruae B. coccirca 65 0 65 0.78 (0.06) 0.76 (0.07) 0.8s (0.04) Section Oncostylß Saies Spicigerac B. ericifolia 51 1 52 0.76 (0.06) 0.73 (0.07) 0.85 (0.0s) B. occidenralß** 74 3 77 0.69 (0.08) 0.65 (0.09) 0.81 (0.04) B. tricuspß **# 54 0 54 0.80 (0.07) 0.78 (0.09) 0.89 (0.0s) Seriæ Dryandroideae B. dryandroides ** 66 0 66 0.74 (0.07) 0.70 (0.09) 0.83 (0.04) SeriæAbieti¡ae B. meßneri tÉ**# 59 0 59 0.73 (0.06) 0.69 (0.08) 0.83 (0.0s) B. pulchella 66 I 67 0.7r (0.07) 0.67 (0.0e) 0.80 (0.07) Subgenus Isostylß B. illicifolia 60 I 61 0.71 (0.08) 0.67 (0.09) 0.80 (0.05) B. oliganthø * 48 0 48 0.73 (0.07) 0.69 (0.09) 0.85 (0.04) B. curuala***# 78 0 78 0.80 (0.09) 0.78 (0.10) 0.90 (0.04) GenxDryandra D, formosa 55 4 59 0.& (0.0e) o.se (0.11) 0.7s (0.0s) D. poþcephela 76 I 't7 0.77 (0.0e) 0.74 (0.11) 0.87 (0.04) D. ca¡li¡toid¿s 59 3 62 0.66 (0.08) 0.61 (0.10) 0.76 (0.0s) Corservation sanrs of Baz,tsia and Dryandra species (George i98Ð. * specieswhicharerarebutnotcurrentlyconsideredendangeredorvunerable ¡l"r' vunerable species not presently endangered but at risk in the longer term *** endangued species which may disappear from the wild within one or two decades if present land use and other causal facton continue # Declared rare flora under tlre Westem Ausralian Wildlife Conservæion Act L95UI979. L29 Genetic diversiw n B anlcsia and Drlandra Table 8.2 Genetic diversity of three species of Banl Specres Allozymes RAPD analysis Nei and Jaccard Russell Apostol Excoffier Rodgers ti and Rao et al. et al. and Tanimoto (re79\ (1901) (1940) (1993) (1992) (1960) B. attenu.ata 1.10 a 0.77 0.74 0.86 o.4t 0.25 o.57 B. menziesií 1.04 a 0.6s 0.59 0.77 0.35 0.29 0.51 B. cuneøta 0.67-0.94 b 0.80 0.78 0.90 0.36 o.28 0.52 a. Scott (1980). b. Coates and Sokolowski (1992). 130 Genetic diveniW n B anlcsia and Dr:r, andra Figure 8.1 Agarose gel showing RAPD markers produced using primer OPB-7 for 10 individuals of B. ashbyi. Lane 1 shows the DNA size standardpGEM (Promega), lanes 2 - 11 show the 10 individuals of B. ashbyi. Bands a¡e scored as present (1) or absent (0), and a matrix of ls and 0s are established for each individual, for all primers used in this study. The matrix is then used in subsequent genetic distance analysis. t3r 12 3 4 5 6 7 8910 11 2645 1605 1_198 61 6 51 46 396 350 ))) I79 Genetic diversity n Barilcsia and Dryandru Díscussíon This study shows that the 33 species of Banlcsta and the three species of. Dryan"dra tested have high levels of genetic diversity. Comparable levels of within species diversity were detected by RAPDs and by enzyme electrophoresis. RAPDs produce large numbers of polymorphic markers, compared to the limited number available for allozyme techniques. In this study RAPD results were reproducible by using optimised standard conditions. Other workers have also shown that by using specific conditions, and taking care to avoid alteration of any of the conditions, reproducible results can be obtained ('Itrk et al. 199s). The relationships between the six distance matrices tested have been discussed by Gower (1935). Some of the metrics a¡e related by simple monotonic functions and the distances they produce are linearly or curvi-linearly related. Thus the metrics of Nei and Li (1979) and facca¡d (1901) form one group, and simple matching (Apostol et aI. 1993), Excoffier et aI., (1992) and Rodgers and Tanimoto (1960) form another, with the metric Russell and Rao (1940) producing distances that are poorly related to those produced by any of the others. This was tested on three data sets using the different metrics, and there was agreemenr between the metrics of Nei and Li (1979) and Jacca¡d (1901), and between simpie matching (Apostol et aI., 1993), Excoffier et aI., (L992) and Rodgers and Tanimoto (1960). The metric of Russell and Rao (1940) agreed with the fìrst group, and with the previously pubtished allozyme results (Scott 1980, Coates and Sokolowski L992). These results are consistent with Gower's (1985) expectations, a¡d suggest that the Nei and Li (1979), Jaccard (1901) or Russell and Rao (1940) methods are the most appropriaæ for RAPD data sets. For the calculation of genetic diversity using RAPD data, it has been assumed that co- migrating fragments are allelic. Homology of cemigrating fragments has been previously demonstrated for different species of G$cine andAllium (V/ilikie et al. l993,Williams er t32 Genetic diversity n Banlcsia and Dryandra al. 1990b), and indirect evidence for allelism is derived from the conformity of taxonomic classifications based on RAPD data to those derived by morphology, cytology and enzyme electrophoresis which are widely accepted (Demeke et al.I99Z,HowelI et al. 1994). While the assumption of allelism is not unrealistic, inheritance of bands in appropriate crosses should be studied wherever possible. Due to the large number of ma¡kers produced and the small amounts of tissue required, RAPDs have potential for use in estimating genetic diversity in populations of rare and endangered species. Estimates of genetic diversity and the distribution of this variation within or between populations of plants is important for devising effective conservation management strategies. RAPDs can also identify plant populations that should be monitored more closely for future conservation. Genetic diversity of all species tested in this study was high, indicating that none has reached dangerously low levels. More research is needed however, to target fragmented species across all remaining populations. In addition, RAPDs can identify individuals within populations, making them ideal for studies of parentage, cultivar identification and geneúc mapping. t33 B. cuneata population genetics Chapter Nine RAPD variation within and between populations of Bønksiø cuneata A.S. George (Proteaceae), a rare and endangered species. Abstract Banl in south western Australia. Random amplified polymorphic DNA (RAPD) analysis was used to determine genetic diversity within and between all of the ten known populations. Estimates of genetic diversity ranged from 0.65 - 0.74, which is high considering.B. cuneata is a rare and endangered species. Analysis of molecula¡ variance (AMOVA) was used to parrition RAPD variation within and between populations. Nea¡ly all of the variation was attributable to individuals within populations, indicating a lack of population divergence. It is suggested that the combination of bird pollination aad high outcrossing rates in B. cuneata maintain genetic diversity and cohesion between the populations. Introduction Banksia cuneata is a ra¡e species known only in 10 populations in the central wheatbelt area of south western Australia, totalling about 550 plants, in an area of about 90 km 2. t is found in deep yellow sands which occupy approximately 10 - 15 percent of this area, giving it a fragmented distribution. Associated with these soils is a rich and diverse flora dominated by species of the famity Proteaceae, Myrtaceae and I-eguminosae. In the last 50 - 60 years land clearing for agriculture and other disturbances have reduced the.B. t34 B. cuneata poPulation genetics cuneata population size to about 77o of its original distribution, and it now occurs as remnants of native vegetation. Honeyeaters are reported to be the major pollinator of B. cuncata and bird populations are supported by co-existing vegetation. Many Banltsia species are pollinated by birds and small mammals, and these pollinators move considerable distances and visit large numbers of plants (Ayre and Whelan 1989). As in the case of many Proteaceae, B. cuneata flowers are protandrous, and outcrossing is likely to be common (Ayre and Whelan 1989). The mating system is an important factor in determining patterns of genetic variation within and between populations (Brown 1989), and estimates of outcrossing in Banksia are typically high (Scott 1980, Carthew et aI. 1988, Sampson er aI.1994). With high levels of outcrossing and bird pollination most genetic variation should be detected within populations, rather than between (Brown 1.979, Hamrick and Godt 1989). Recently; the mating system and patterns of genetic va¡iation for 6 populations of B. cuneata were determined using enzyme electrophoresis (Coates and Sokolowski t992). Estimates of outcrossing based on 6 loci, ranged from 0.67 - 0.95, with low levels of selfing. The populations were divided into two groups, with gene flow within groups, but not between. It was suggested that an ecological barrier to the pollinators may be responsible for the restriction of gene flow benveen groups. Banksia species along with many other members of the Proteaceae, show low levels of allozyme variability (Whelan 1994). More recently, random amplifiedpolymorphic DNA (RAPD) analysis based on PCR technology (Williams et aI.1990) has been used to address variability within and between poputations of a range of genera (Fluff et al. L993, Chalmers et at. 1992, Russell et aI. 1993). B¡r--using-+ingle+andorn-priurers.-this' +eohn.ique-can amplify regions wi.thin the genome,-'without**any-pd€r--€equen€e- in$snnatiogusing small. amounts of tissue. It is also .roasonably- sirnple and-e-an-prorride-a veryJar.ge-nurnber of potential poly,rnorphio loci TÏris-is-ideal-in the consorvatisncontext' æfua¡e-a¡rd-endangelodçlant+;sineethere-is-ofterÈfittl€-ma8edd-and-lew-varhbi+ity- 135 B. cuneata PoPulation genetics deæe+cd-"ì¡s.ing*allozyrnes, The aim of this study is to determine patterns of genetic va¡iation within and between all known populations of .8. cuneata using RAPDs. Four populations are included in this study which have not been investigated previously. This information is important for devising appropriate conservation strategies for the remaining B. cuneata populations. Materials and methods Population sampling Seed material for RAPD analysis was collected from 6 populations by the Department of Conservation and Land Management W.A. Additional seed was collected during a field trip to Western Australia (Fig 9.1). Twenty seeds were randomly selected, one per plant, from four populations with in excess of 50 plants, and ten seeds were collected, one per plant, from 3 populations with less than 50 plants. In three cases, the populations comprised less than 10 plants, so samples of one seed were taken from all remaining plants. In total 125 plants, approximately 25Vo, of the total number of plants remaining were sampled. DNA isolatíon DNA was extracted from each seed using a slight modification of the method of Weining and Langridge (1991), comprising: phenofchloroform/isopropanol extraction for 5 min on ice, DNA precipitation for 1 min with ice cold isopropanol and sodium acetate, with DNA recovered by centrifugation at 12 000 rpm for 10 min. The pellet was washed twice with 70 Vo ethanol, dried and dissolved in 50 pL of TE buffer, with 1.0 pL of RNAase (R40: 40 glmL RNAase A in TE), and stored at 4 0C short term (up to 1 month) or -20 oC long term (up to 1 year). r3ó B. cuneatapoPulation genetics DNA was subjected to gel electrophoresis oî a L.67o agarose gel in TBE buffer (Sambrook et a\.1989), and stained with ethidium bromide. DNA concentration was estimated by visual assessment of band intensities, compared to salmon sperm genomic DNA standa¡ds. The DNA content was adjusted to 10 ng pL -1. DNA amphrtcafion and docuttæntation DNA amplification was performed in a MJ Research Thermal Cycler. The optimal program for Banksía commenced with an initiat denaturation step at94oC for 5 min, followed by 40 cycles of 94 0C for 1 min, 36 0C for 1 min, 720C for 2 mins, terminated with a final extension step at720C fo¡ 5 min. Optimised reaction conditions were ca¡ried outin a25 ¡tL total volume containingLXTaq buffer (Gibco-BRL),3 mM MgCl2,200 pM of each dNTP (dGTP, dATP, dCTP, dTAP), 1 unit of. Taq polymerase (Gibco- BRL), 0.5 pL T4 gene 32 protein (Boehringher Mannheim), 1 pM 10 mer primer (Operon Technologies Inc.) and 10 ng pL -1 template DNA. Each reaction mix was overlaid with PCR grade paraffin oil. DNA amptification fragments were separated by 27o agarose gel (Seakem, Promega) electrophoresis using TBE buffer (Sambrook et al. 1989). A DNA size marker was used (pGEM, Promega) on each gel, and gels were stained with ethidium bromide. Fragment patterns were photographed under UV light with Pola¡oid 667 frlm for further analysis. Polaroid photographs were scanned using a transmission scanner (Hewlett Packard Scanjet IIcxfI). The intensity and molecula¡ weight of each band was determined using the software CREAM TM., Kem-En-Tec Softwa¡e Systems, Blue Sky Scientifrc. Sixty primers were evaluated for thei¡ suitability in a pilot survey (series OPA, OPB and OPC, Operon Technologies Inc.). Five primers were selected (OPA-1, OPA-4, OPA-9, OPA-I1, OPA - 16) that gave reproducible and informative markers. Band fragments included in the frnal analysis ranged between 2.5 kb and 100 bp in length (Fig 9.2) and were scorcd as present (1) or absent (0). A negative contol was added in each run to test L'J'I B. cuneata population genetics for contamination. In order to test reproducibility, the selected primers were tested three times on the same sample, for a random subset of three DNA samples. To aid interpretation of band homology between gels, each gel contained a standard DNA sample and pGEM DNA marker. The presence or absence of bands was determined for all individuals and a matrix of RAPD phenotypes was assembled. Each individual was represented by a vector of 1s and 0s for the presence or absence of any particular band for all the primers used in the study. Sntístical analysis The vector of presence/absence states for each individual was used to compute a measure of genetic distance for all pairs of individuals using the method of Nei and Li (1979). It was determined previously that the metric of Nei and Li was one of the most suitable for RAPD data on Banksia species, compared to known diversity estimates based on allozyme data (chapter eight). The index of genetic distance was calculated as D = 1-F. 'Where F = similarity (Nei and Li 1979), calculated by the equation 2*nlll((2*n11)+n01+n10), where n = number of band positions, nl1 = the number of posiúons where x = 1, y = 1, n00 = the numberof positions where x = 0, y = 0, n01 = the number of positions where x = 0 and y = 1, nLO = the number of positions where x = 1 and y = 0, x = individual number x, and y = individual number y. Distances were calculated using the statistical package, RAPDistance (Amstrong et al. 1994). Following analysis of distance within each population, a mean estimate of genetic diversity and standard deviation was derived for each population. Analysis of molecular variance (AMOVA) (Excoffier et aI.1992) was used to estimate variance components attributable to differences within and between populations. Significance levels for variance component estimates were calculated by permutational procedures. The number of permutations for significance testing was set at 100 for all 13E B. cuneata population genetics analyses. The analyses were undertaken with the AMOVA program for'Windows (MNAMOVA) provided by L. Excoffier. An unweighted pair group mean average (UPGUA) clustering analysis was carried out using a distance matrix based on the Phi sratisric (PhiST) produced by the V/INAMOVA analysis for between population distances. Results The RAPD profile After excluding bands that were greater than 2.5 kb or less than 100 bp for the whole data set, the five primers yielded a total of 169 polymo¡phic bands (Table 9.1). The number of ma¡kers per primer ranged from 30 (OPA - 4) to 37 (OPA - 9, OPA - 16). Of the 169 RAPD markers scored, 33 (I9.5Vo) were present in all populations and there were no fixed differences between populations, with the remaining 136 (80.57o) bands variable between populations. Most bands occurred at a frequency of 0 - 207o, with relatively few bands in the 80 - l00%o category (Tabte 9.2).The total number of bands present in each populaúon was similar, with the exception of population 10, which had nearly half the number of bands (Table 9.2). Thus, for populations with between 6 and 20 individuals there was a relatively high number of bands, but population 10 with only 3 individuals, had 66 bands. Estimate of genetic diversiry The estimate of genetic diversity (calculated as D = 1 - F) for each population of B. cuneata, and the whole species is shown in Table 9.3. The total genetic diversity for ^8. cuneata was 0.70, and the diversity within each population ranged from 0.65 þopulation 10) to 0.74 þopulation 2). The highest genetic diversity estimates occurred in the larger populations (population2,3,4, 9), with less diversity in smaller populations (1, 5, 6,7, L'J9 B. cwteata population genetics 8, 10). However, genetic diversity within a population did not always depend on population size; population 6 had 6 individuals with a diversity of 0.70, which was higher than populations 1, 5, and 7 which had 10 individuals. Results of the AMOVA analysis are shown in Table 9.4, which attributes nearly all of the variation to individuals within populations. Analysis between populations shows a small negative variance component, indicating lack of population structure at this level, so some plants between populations are more related than within populations (Excoffier et aI;": lgg2). Therefore, all variation is attributable to individuals within populations and between populations (ie. among all individuals as a group), for which the corrected percentage of variation due to individuals is about L00Vo (Table 9.4). These results suggest that there is no distinction be¡veen populations of B. cuneata. This is reflected by the absence of fixed differences between populations in RAPD markers. To show relationships between the populations (Figure 9.3), UPGMA clustering was applied to the distance marrix calculated by the AMOVA analysis, based on tho Phi statistic (PhiST) metric @xcoffier et al.1992). de,r1+"r"'.o* From the t$läoâùm-some interesting relationships arise; geographically very close populations such as 6, 7, and 3 are in one cluster, closely related to populations 2, 4 and 10. Another clusrer contains populations 5, 8 and 9, with population 1 linking the two clusters. Interestingly, population 9 and 10 are the most southerly populations, with one population in each cluster group. This clustering, with low levels of genetic diversity between populations, reflects the lack of genetic differentiation between all ten populations. 140 B. cuneatø PoPulation genetics Table 9.1 Summary of data obtained by RAPD analysis for 5 primers with 125 individuals of B. cuneata.The table shows the sequence, total band number, number of polymorphic and monomorphic bands, and the average number of bands/primer. Number of bands Primer Primer sequence Total Polymorphic Monomorphic 5' to 3' OPA- 1 CAGGCCCTTC 33 33 0 OPA-4 AATCGGGCTG 30 30 0 OPA-9 GGGTAACGCC 37 37 0 oPA - 11 CAATCGCCGT 32 32 0 oPA - 16 AGCCAGCGAA 37 37 0 Total r69 t69 0 Mean perprimer 33.8 33.8 0 14r B. ctneata PoPulation genetics Table 9.2 Summary of band frequencies for each population of B. cuneata. The table shows the number of bands out of 169 (total data set), for each population that fall into frequency groups of:0 - 20Vo,20 - 407o,40 - 60Vo,60-807o, and 80 - 1007o. Band frequency Population Number Number 0-207o 20-407o 4O-60Vo 60-8OVo 80- lOÙVo of plants of bands sampled 1 10 13t 91 28 23 15 t2 2 20 L57 93 37 19 L3 7 5 20 161 78 43 24 18 6 4 20 145 83 47 20 11 8 5 10 128 90 28 24 T6 11 6 6 115 94 29 11 18 t7 7 10 135 96 24 23 T2 T4 8 6 118 90 20 23 l4 22 9 20 t45 9T ?) 23 18 5 10 J 66 103 34 0 15 t7 Mean 130.1 90.9 32.2 19.0 15.0 11.9 T42 B. cuneata population genetics Table 9.3 Genetic diversity of each population of. B. cuneata catculated using the simitarity (F) metric of Nei and Li (1979), where distance (D) = 1 - F. The values shown are the mean distance (diversity) values for each population, and the standard deviation of the mean. Population Number of Mean distance Standard plants sampled (diversity) deviation 1 10 0.68 0.085 2 20 0.74 0.096 5 20 0.73 0.067 4 20 0.73 0.053 5 10 0.68 0.073 6 6 0.70 0.065 7 10 0.67 0.070 8 6 0.67 0.032 9 20 0.71 0.080 10 J 0.6s 0.017 Total t25 0.70 0.031 L4'3 B. cuneata PoPulation genetics Tabte 9.4 Analysis of molecular va¡iance (AMOVA) for 125 individuals of. B - cuneata using 169 RAPD bands. The 125 samples a¡e divided into 10 populations. The data show the degrees of freedom (dÐ, sum of squared deviation (SSD), mean squared deviation (MSD), variance component estimate, percentage of toÞl variance contributed by each component and the probability of obtaining a more extreme component estimate by chance alone. Source of variation df SSD MSD Variance Vo total P - value component ArøIy s is ømo ng p opulatio ns Between populations 9 1.446 0.161 -0.0160 -4.73 7o 0.57 Within populations 115 40.821 0.355 0.3s49 104.73 Vo < 0.01 IM B . ctneata oootúz!¡on genetics Figure 9.1 Geographic range oL Banksia cuneata in south-west Australia, showing locations of the ten populations from which seed was collected. r45 \\/ESTERN A iJSTRALIA 0 300 km /-\( I PERTH ,ú Quairading Bo \ 7 0t6 3 e 30 kr¡ lro I I ! I PERTI{ .l B. cuneata population genetics Figure 9.2 Agarose gel showing RAPD markers produced using primer OPA - 9 for individuals of B. cuncata.Lane L shows the DNA size standard pGEM @omega), lane2 shows the standard individual to aid in band homology for between gel comparisons, lanes 3 - 6 show individuals in population 7, lanes 7 - 16 show individuals in population 3 and lane 17 is the negative control. Bands within the molecular size range 2.5 kb to 100 bp are scored as present (1) or absent (0), and a matrix of 1 s and 0 s are established for each individual, for all primers used in this study. The matrix is then used in subsequent genetic distance and AMOVA analyses. r4ó 12 3 4 5 6 7 A 910'11 12 13 14 15 1617 2645 160s 119 676 51 46 396 350 222 1_1 9 B. cuneata populiation genetics Figure 9.3 UPGMA cluster analysis of B. cuneata poptlations based on genetic distance (PhiSÐ among populations calculated by AMOVA analysis. t47 Genetic distance 0 . or2'7 0.0232 0.033? 0.0443 0.0548 0.0653 I I I Populacion 6 Popu J-ac ion 2 Popu 1a t ion 1 Populat ion 3 Population 10 I Popu Iat ion 4 Population 1 PopuJ-ation 5 I Populatj.on 9 II PopuJ-at ion 8 I_l I I I I I .t o.or2'7 0 .0232 0.0337 0 0443 0.0548 0 n 6s3 B. cuneato DoDulation genetics Discussion Banksía species a¡e considered to be bird and mammal pollinated, and have high levels of diversity.wittr outcrossing rates amongst the highest recorded for plants (Scott 1980' Carthew et al. L988, Sampson et aI. L994). Genetic diversity in B. cuneata as a whole estimated by this study is quite high, considering it is a geographically restricted rare and endangered species. The 10 known populations of B. cuneata show differences in the extent of disturbance, and two populations (9 and 10) have a southerly distribution. Populations 1, 6, 7, and 10 occur in highty disturbed roadside vegetation, populations 3, 4, 8 and 9 show low levels of disturbance, while populations 2 and 5 occur in virnrally undisturbed vegetation. There were differences in genetic diversity calculated for each population, population 10 a highly disturbed small population of only 3 individuals, had the lowest genetic diversity of 0.65, while some of the other smaller populations had estimates ranging from 0.67 - 0.70. The larger populations with low levels of disturbance, had the highest diversity estimates ranging from 0.71 - 0.74. Variation in genetic diversity estimates between populations is not uncommon in species with mixed mating systems (Schoen 1982). Pollinator availability and activity are likety to have an influence on genetic diversity berween populations. This particularly applies to bird pollinated species such as B. cuneata. Population structue such as plant density, and level of disturbance may also influence pollinator behaviour. This is reflected by larger populations with lower levels of disturbance, having high levels of genetic diversity. In contrast, small populations on road verges with high levels of disturbance, display lower levels of genetic diversity. The estimates of genetic diversity calculated in this study are in agreemont with enzyme electrophoresis (Coates and Sokolowski 1992) methods which estimated outcrossing rates forB. cuneata.,ranging from 0.67 - 0.95, with low levels of selfing. t4ð B - cuneata oooulation genetics Dissectedpopulation structue and geographic restriction of B. curßata would be likely to influence the patterns of genetic variation within and between populations (Hamrick and Godt 1989). As B. cuneata is outcrossing and bird pollinated, most variation would be expected. to be within populaúons and little between populations. The results of the AMOVA ciearly demonstrate this, with nearly all variation attributable to individuals within populations, with no significant genetic differentiation between populations. AMOVA is not strictly rigorous with non Euclidean metrics, but as ExcofFrer et aI. (1992) point out rhe metric of Nei and Li (1979) differs only in the choice of denominator and the metric is nearly interchangeable with the Euclidean metric. The AMOVA method, designed for other molecular data, has recently been used to analyse RAPD data for buffalograss (Huff ¿r al. L993), Eucalyptus (Nesbitt et al. L995) andGrevil/eø (Rossetto et al.1995) The lack of population differentiation in B. cunearø using RAPDs contrasts to a previous study by Coates and Sokolowski (1992), using six of the ten populations of B. cuneata, and enzyme electrophoresis. Their study using 6 potymorphic loci, found significant differentiation of populations into two groups. It was further suggested that differentiation may be due to a salt river system acting as an ecological barrier to pollinator movement. The westerly populations in particular, were more heterogenous and it was found that some populations which were geographically closer, were moro related to other populations. Their study also found that in the eastern populations, gene flow was relatively high, suggesting that birds as pollinators were effective in maintaining genetic cohesion and diversity benveen the populations. The results of the present study using RAPDs on all 10 known populations show no signif,rcant population differentiation, with all variation between individuals. This suggests that outcrossing and bird pollination are very effective in maintaining genetic diversity and cohesion between populations. Birds are able to travel conside¡able I49 B. cuneata oopulation genetics distances between populations, and associated with areas of B cuneata is a rich and diverse vegetation system which can maintain bird populations. Since clearing for agriculture, which was about 50 - 60 years ago, there has been no signifrcant differentiation between the populations, although population sizes have reduced. This could be for a number of reasons. It is possible that after clearing the populations are still visited by pollinators due to other surrounding vegetation, which support the bird population. It could also be because the populations are declining and environmental conditions are generally unfavourable to seedling recruitment, that the populations a¡e agrng and there has been few generations after clearing to see any observable amount of genetic differentiation between populations. Furthermore, it is possible that prior to clearing there were no significant ecological ba:riers to pollinator movement and the ecological banier suggested by Coates and Sokolowski (1992) is relatively recent, therefore no significant divergence of the populations has occurred- It is important for future conservation and persistence of B. cuneata that the remaining populations arc protected. Even though they have reasonably high levels of diversity, this study shows in very small populations, such as population 10 with only 3 individuals, that genetic diversity may decline. Pollinator maintenance is essential for maintaining genetic diversity, and reserves which have other coexisting vegetation to support pollinators must also be protected. The distribution of these reserves to allow gene flow between populations, is important for the survival of. B. cuneat¿, as well as maintaining genetic diversity in other plant species. Traditionally, studies of population genetic structure have used allozyme markers (Hamrick and Godt 1989). Allozymes can provide informative markers for levels of generic variability within a plant species. It has been shown that genetic diversity is generally greater at the population and species level for outcrossers than selfers. In addition, widespread species exhibit more variation than restricted species (Ilamrick and Godt 1989). There are however, widely recognised limitations of allozymes. Many 150 B. cuneata DoDulation genetics species of the family Proteaceae have low levels of detected allozyme diversity. This is because the detection of variability is limited to protein coding loci, which may underestimate levels of genetic diversity (Clegg 1989), and may not represent the entire genome (Schaal et aI.l99l). Allozymes are also tissue specific and are influenced by environmental factors. Given these limitations many workers have moved to using DNA markers such as RAPDs. They overcome the limitations of allozymes and are simpler and easier to use than other DNA markers such as restriction fragment length polymorphisms (RFLPs). RAPDs have potentially unlimited numbers of markers and have been shown to be useful for a variety of applications including population genetic studies on a number of genera (Huff et a\.1993, Chalmers et aI.I992,Russell et aI.1993). Direct comparisons of RAPD and allozyme diversity are few. Peakall et al., (1995) compared RAPD ma¡kers to allozyme markers in buffalograss. Using AMOVA analysis, they found that both markers gave similar qualitative patterns,large regional differences and significant population differentiation within regions for four populations of buffallograss. RAPDs however, detected more variation than allozymes. Dawson et al., (1993) found patrerns of RAPD variation \n Hordeum spontaneum comparable to previously reported patterns using allozymes. Liu and Furnier (1993) using two species of. Populus compared RAPD, RFLP and allozyme markers. They found that RFLP and allozyme markers gave comparable patterns within and between species, while RAPDs revealed more variation between individuals within a species than allozyme and RFLP markers, and produced reliable discrimination among clones. Our findings with B. cuneata are consistent with reports that RAPDs detect more variation than allozymes (Peakall et a1.,1995, Liu and Furnier L993, Russell et al.,1993, Dawson et al., 1993). Furthermore, our study using RAPDs shows nearly all va¡iation is within populaúons of B. cuneata rather than between populations. Russell et al., (1993) also reported RAPD diversity in Theobroma cacao was higher within provinces than between provinces, a pattern typicat for outcrossing woody plants (Hamrick and Godt 1989). Although allozymes a¡e valuable for many typos of studies, it is clear that DNA markers such as 15I B. cuneata Dopulation genetics RAPDs will be widely used on plant population studies in the future, and will remain a valuable genetic tool. In conclusion, Banksía cuneata has relatively high levels of genetic diversity detected using RAPDs, with no significant differentiation between populaúons over a restricted geographic range. Gene flow is high between populations, apparently due to pollinator movement and a predominantly outbreeding mating system. Future conservation of the species will depend on adequate protection of the remaining populations, collection of seed material from populations into a gene bank, and maintenance of bird populations in surrounding and co-existing vegetation areas. L52 BanksiaRAPD phvloseny Chapter Ten Use of RAPD markers to analyse phytogenetic relationships in Banksía (Proteaceae) Abstract Random amplifred polymorphic DNA (RAPD) analysis was applied to 33 species of Banlcsia, three species of Dryandra andMwgravea heterophyll¿ in order to investigate phylogenetic relationships within genus Banksia, and between subgenera and genora. Fifteen primers produced 791 bands for the 37 taxa, and results were analysed using group average (UPGMA) clustering and parsimony (PAUP). The resulting analyses were in broad agreement with each other, with closely related species pairs and species groups clustering together, as in the accepted classification of Banksia. RAPDs were not informative between distantly related species or species groups. Introduction Early treatments of Banksía were either catalogues of species (Brown 1810), or artificial classifications (Meissner 1856). George (19S1) attempted a moro natural classification, using a variety of leaf, flower, follicle and seed characters to base infrageneric taxa. While George (1981) describes "possible lines of evolution" within and bet'ween tæra, his classification is difficult to interpret phylogeneticalty because of a number of taxa with uncertain relationships. Doust (1983) attempted a cladistic analysis of. Banlcsia using 37 characters for 86 taxa, but the trees were not clearly resolved. The analysis provided a poor march with the existing taxonomy and faited to suggest a robust phylogenetic scheme. It was not used for a formal classification. More recently Thiele (1993) proposed a classification based on cladistic analysis of morphological characters. Although there is r53 B anl -Ac¡e*+¡¿-eomprises¡¡vo-sub€sner"q-Jsas ryJis-ædBanksi+{ees€æ group-immal-l-and-the-spee{erar+*uperfreia"lly+-im.ilar-to-Dryandrø-'This-Ied*George GgSl)Jo+roposc-th^at"Isas4,J¿sÆay-be-close*tø-D.ryørù-a-thaa-te-sì¡bgen¡rs+.e#H4' Neve*tlreleq*heplaeed-it-as-a-subge'nu*-of-Bø#s¿:ø*vitlrtheeomrnen#æeparate- €€,nüffia.f be-appr.opriate, Thi-s-ìnærprot+tior+-was -base&-en-+h@ js-stilL.speeulati,on--on €hæect€r$,Jhere - tåe".rola-tionshþ s-wi+hin@ mbgenrrs-ftosA¡¡ts, subgenus-Bør*srø*n@ Other data such as DNA markers may help to resolve these relationships. The development of random amplified polymorphic DNA (RAPD) markers generated by the polymerase chain reaction (PCR) using arbitrary primers has resulted in molecular markers for the detection of nuclear DNA polymorphisms (Williams ¿r al., 1990). The technical simplicity of the RAPD technique has facilitated its use in the analysis of phylogenetic relationships in several genera (Wilkie et a1.,1993, Demeke et al., 1992, Abo-elwafa et aI., 1995). In this study RAPDs were evaluated as informative markers at different taxonomic levels within Banksia. In addition, this approach was used to investigate higher order relationships between related subgenera and genera. r)4 BanksiaRAPD phvlosenv Materials and methods Planr marcríal Seeds of 33 species of Banksía and three of Dryandra, collected from natural populations, were obtained. from Nindethana Seed Service. Fresh leaf material of Musgravea heterophylla was collected from Atherton, Queensland (QRS Arboretum No. 1372). The species chosen represented three genera Musgravea, Banlcsia and Dryandra, and two subgenera, two sections and thi¡teen series within genus Banksia (George 1981, 1988) (Table L0.2). Ten seeds of each species of Banksia and Dryandra were randomly selected for RAPD analysis, and the DNA bulked to give one sample per species. One leaf was selected îor Musgravea luterophylla. DNA extractíon DNA was extracted from each seed using a modified method of \Veining and Langridge (1991), comprising: phenoVchloroform/isopropanol extraction for 5 min on ice, DNA precipitation for 1 min with ice cold isopropanol and sodium acetate, with DNA recove¡ed by centrifugation at 12 000 rpm for 10 min. The pellet was washed twice with 70 7o ethanol, dried and dissolved in 50 pL of TE buffer, with 1.0 pL of RNAase (R40: 0C 40 glmLRNAase A in TE), and stored at 4 0C short term (up to 1 month) or -20 long term (up to 1 year). DNA was extracted from fresh leaf material of Musgravea heterophylla using ttre extraction method for seedling material of Maguire et aL (1994). DNA was subjected to gel electrophoresis on a 1.67o agarose gel in TBE buffer (Sambrook et aI.1989), and stained with ethidium bromide. DNA concentration was estimated by visual assessment of band intensities, compared to salmon sperm genomic DNA standards. The DNA content was adjusted to 10 ng FI -1. 155 BanksiaRAPD phylogeny DN A amplifi c atí o n and do cumc nt atio n DNA amplification was performed in a MJ Research Thermal Cycler. The optimal progam .for Banksia commenced with an initial denaturation step atg4oc for 5 min, followed by 40 cycles of 94 0C for 1 min, 36 0C for I min, 720C for 2 mins, terminated with a final extension step aú2ocfor 5 min. Optimisedreaction conditions were carried 200 out in a25 ¡tJ- total volume containin g lX Taq buffer (Gibco-BRL), 3 mM MgCl2' pM of each dNTP (dGTP, dATP, dCTP, dTAP), 1 unit of Taq polymerase (Gibco- BRL), 0.5 pL T4 gene 32 protein (Boehringher Mannheim), 1 pM 10 mer primer (Operon Technologies Inc.) and 10 ng pL -1 template DNA. Each reaction mix was overlaid with PCR grade paraffin oiI. DNA amplification fragments were separated by 2Vo agarose gel (Seakem, Promega) electrophoresis using TBE buffer (Sambrook er a/. 1989). A DNA size marker was used (pGEM, komega) on each gel, and gels were stained with ethidium bromide. Fragment patterns were photographed under UV light with Polaroid 667 film for further analysis. Polaroid photographs were scanned using a transmission scanner (Hewlett Packa¡d Scanjet IIcxÆ). The intensity and molecular weight of each visible band was determined using the software CREAM rM., Kem-En- Tec Software Systems, Blue Sky Scientific. Sixty primers were evaluated for their suitability in a pilot survey (series OPA, OPB and OPC, Operon Technologies Inc.). Fifteen were selected (Table 10.1) which gave reproducible and informative markers. Band fragments were scored as present (1) or absent (0) Grg 10.1). A negative control was added in each run to test for contanination. In order to test reproducibility, the selected primers were tested three times on the same sample, for a random subset of three DNA samples. To aid interpretation of band homology between gels, each gel section contained a DNA marker (pGEM). The presence or absence of bands was determined for all individuals and a matrix of RAPD phenoq?es was assembled. Each individual was represented by a vector of ls and 0s for the presence or absence of any particular band for all the primers used in the study. 15ó BqnlniaRAPD phylogeny Dan aralysis Two alternative methods were used to analyse the RAPD data. A genetic distance matrix (Nei and. Li t979), with group average (UPGMA) clustering was used to produce a d e nd loo'lad\ æn¿eþa*. The second method applied phylogenetic analysis using the parsimony program PAUP (Swofford 1990). G ene ti c dí s tanc e analy s ís From the matrix of RAPD phenotypes for each species, an index of genetic distance was calculated @ = 1 - F), where F is similarity calculated using Nei and Li (1979) matching coeffîcient merhod [2*nl ll((2*nl1)+n01+n10)]. Where n=number of band positions, nl1= the numberof positions where x=l, y=1, n00= the number of positions where x=0, y=0, n01= the number of positions where x=0 and Y=1, n10= the number of positions where x=1 and y=0, and x,y = individuals being compared. The distance matrix was calculated using the statistical package, GENSTAT 5 (Payne 1987), and UPGMA clustering was performed by the PATN program package (Belbin 1991). Phylogenetíc analysis using parsímony (PAUP) RAPD data obtained from the 37 taxa, with Musgravea as the outgroup, were used to construct phylogenetic trees. Analysis was performed using PAUP 3.1.1 (Swofford 1990). The phylogeny was assessed using the heuristic search method of PAUP (characteroptimisation ACCTRAN, MULPARS andTBR branch swapping options). To ensure that all islands of most parsimonious trees were found (Maddison 1991), the search was repeated 100 times with RANDOM addition and with a maximum of 100 trees saved at each replication. The resulting most parsimonious trees produced by this process were subjected to successive weighting to minimise homoplasy. The matrix was ß1 BanluiaRAPD phyloseny reweighted until the tree-length values stabilised. Equally short solutions were used to produce a consensus tree. Results A total of 791, bands 'P"iascored, with an average of 52.7 bands per primer (Table 10.1). The banding pattern for 34 species in the study using primer OPB-11 is shown (Fig 10.1). The number of polymorphic markers was high, with few monomorphic bands, and the number of bands unique to only one species comprised 18.5 Vo of the toral. All species generated similar numbers of bands per primer (Table 10.2), except for B - cuneata, B . olígantla and M. heterophylla which produced more than average. In cluster analysis of genetic distance values, Musgravea formed a small group with Banksia, subgenus Isosrylis, and the three species of Dryandra formed a small group with Banksia coccine¿. For the remainder of genus Banksia, species pairs and closely related species generally formed groups @g 10.2). The cluster involving Mu'sgravea and subgenus Isosrylis is interesting, as this group is closer to many Banksia species, than they are to each other. UPGMA clustering forms g¡oups with increasing genetic distance. devlcl coqcaon ThedendegËe*'shows approximately 7 - 8 main groups, linked together at gonetic distance values ranging from 0.70 - 0.78. The higher levels of clustering between these main groups is tenuous, however clustering within these groups is likely to be real. Within the main clusters, species pairs and closely related species group together, such as B. occidentalis andB. meisnerí,D. polycephala, D. carlínoides andD. formosa,B. ericiþIia and B. dryandroides, B. caleyi and B. lemanniana, B. grandis and B. solandrí, B. audax and .8. lindleyana, B. oligantha and B. cuneata, and B. ashbyi and B. eld¿riata. The PAUP program identified two minimum length trees (length = 4079) with a consistency index of 0.929. Both ttre stict consensus troe and the majority m1e tree show 15E B anlcsia RAPD phyloseny that species within series Grandes andTetagonae grotrped together, and closely related species between series grouped together (Fig 10.3; 10.4). There were close relationships between species in series Banksia and Crocinae, series Dryandroídeae, Spicigerae and Abíetinae, and series Bauerinae and CyrtosryIts. Dryandra was polyphyletic within Banksia. D. polycephala and D. carltnoídes, as sister taxa, were part of a clade with Banl 159 B anlcsia RAPD phylogeny Table 10.1 Primer sequence and band data for each of the 15 selected primers over the complete data set of 37 taxaproducing 791 bands. Primer Sequence 5'-3' Total bands Polymorphic Monomorphic Unique OPA-1 CAGGCCCTTC 49 49 0 7 OPA-3 AGTCAGCCAC 56 56 0 7 OPA-5 AGGGGTCTTG 55 55 0 9 OPA-7 GAAACGGGTG 53 53 0 6 oPA-13 CAGCACCCAC 53 53 0 9 oPA-20 GTTGCGATCC 51 51 0 15 OPB-1 GTTTCGCTCC 52 52 0 T4 OPB-3 CATCCCCCTG 54 54 0 L3 OPB-4 GGACTGGAGT 50 50 0 9 OPB-6 TGCTCTGCCC 45 45 0 9 OPB-7 GGTGACGCAG 50 50 0 L2 oPB-10 CTGCTGGGAC 60 60 0 8 oPB-l1 GTAGACCCGT 54 54 0 11 oPB-17 AGGGAACGAG 49 49 0 11 OPC-2 GTGAGGCGTC 60 60 0 7 Total bands 79r 79t 0 t47 Mean oerorimer 52.7 52.7 0.0 9.8 160 Bankiø RAPD phylogeny Table 10.2 Band data for 33 species of. Banksía, three of Dryandra and Musgravea heterophylla usin g RAPDS. Subgenus.Bcntsia Section Bantsia Settes Salicina¿ B. integrifolia 138 9.2 B. robur Senes Grandes B. granlis r33 8.9 B. solanl¡i 173 11.5 SaiesQuercinac B. querciþlia r45 9.7 Series Bau¿rina¿ B. bøteri 119 7.9 Series B¿n*sia B. baxteri r42 9.5 B. candollea¡a r54 10.3 B. meraiesii 135 9.0 B. serrata 103 6.9 Series Crocinae B. burdeuii t28 8.5 B. priorøtes r63 10.9 Series Cyrrosrylis B. ashbyi 131 8.7 B. attetaata 134 8.9 B. audax 128 8.5 B. eldcriana 135 9.0 B. laevigata t4l 9.4 B. lindetyarc 150 10.0 B. praemorsa 125 8.3 Seies Prostratac B.blechniþlia 135 9.0 B. repens 118 7.9 SenesTetragorae B. caleyi 113 7.5 B.lemarniana 119 't.9 SenesCoccincac B. coccirwa t37 9.1 Section Orcostylß Seuies Spicigerae B. ericifolia L22 8.1 B. occidentalis 108 7.2 B. tricuspis r20 8.0 Senes Dryandroidcac B. dryandroides 147 9.8 Ser:eß AbietinaÊ B. meßruri t39 9.3 B. pulclwlla T& 10.9 Subgenus Isostylis B. illicdolia r33 8.9 B. oligantha 223 14.9 B. cun¿ata 237 15.8 GenusDryandra D.formosa 158 r0.5 D. polycephela r54 10.3 D. carlinoides 154 10.3 GerusMusgrattea M. heterophylla 240 16.0 161 B anl Figure 10.1 Agarose gel showing RAPD markers produced using primer OPB - 11 for bulked DNA samples of 34 species. Lane 1 shows the DNA size standard pGEM (Promega), lane 2 B. ashbyi, lane 3 B. elderíana,lane 4 B. menzíesii,lane 5 B. blechniþlia,Iane 6 B. integriþIia,lane 7 B. quercfolia,Iane 8 B. Iaevigata,Iane9 B. attenuata,lane 10 B. líndleyana,Lane LI B. repens, Iane L2 B. ílliciþIia,lane 13 B. baueri,lane 14 B. audax,lane 15 B. meisneri,Iane 16 B. praenxorsa, lane 17 B. occidentals, lane 18 B. robur,lane 19 DNA size standard pGEM (Promega), lane 20 B. burdettii,lane2l B. ericiþIia,lane22 B. serrata,lane2S B. dryandroideJ, lane 24 B. candolleana, lane 25 B. príonotes,Iane26 B. batcteri,Lane2T B.Iemanniana,lane 28 B. caleyí,Iane29 B. tricuspis, lane 30 B. grandis,lane 3I B. solandri,lane32 B. pulchella, lane 33 B . coccinea. lane 34 D. formosa,lane 35 D . carlinoídes, lane 36 D. polycephala. Bands a¡e scored as present (1) or absent (0), and a matrix of 1 s and 0 s are established for each species, for 15 primers. The matrix is then used in subsequent genetic distance and PAUP analyses. t62 1 2 3 4 5 67 I 91011'12131415161718 2645 1605 1198 --f- o/ -ta-- ¡ 51 460 - 396 - 35 a¡- tl9 2645 5 1605 - 119 I rú -_-- 6'7 6 -_ ú 517 -i 460 396 350 222 179 1920 21 2223242526 27 28 293031 323334 35 36 BanksiaRAPD phyloqeny Detr¿\roqra\an Figure 10.2 eends$e+generated by cluster (UPGMA) analysis of genetic distance values generated from 791 RAPD bands using 15 primers. Genetic distance @) was calculated as D = 1-F, where F= similarity calculated using the method of Nei and Li (r97e). 1ó3 Genetic dist.ance 0. s330 0.5814 o .6298 0 .6't 82 0.1266 0 ??s0 I I B ashbvi B el-deriena B menzìesìr B blechn:.foIi-a B incegri íol-ia B J-aeviga t a l- B quercifolia T M heE.e roph yI Ia I B ol- igantha I B cuneata I I B attenua E a B rePens B illicifolia I B J-indleyana I D audax I B baueri I B c ricuspis Þ puLchel La -t I B gra ndis I I B solandri I B burdett ii T B cando Il-eana I B prionot.es I I B baxte ri I B lemanniana I I B cal-eyi I B serrat a I I B ericifol-ia I B dryandro ides I B coccrnea T D forrnosa I D ca rl- inoides I D pol-ychephyla I B me)-snerr B occidenE a I i-s I B Praemorsa I B robu r I I I I I I 't1 0.5330 0.5814 0.6298 0 .67 82 c.1266 0 5 0 B anl Figure 10.3 Majority rule consensus tree obtained from PAUP, tree length = 4079; consistency index = O.929. Taxa represent t'wo closely related genera Banksia and Dryandra, with Musgravea as the outgroup in the analysis. Tree generated from 791 RAPD bands produced with 15 selected primers. tó4 B.ashbyi B.elderiana B.menzeisii B.blechnifolia B.integrifolia B.quercifolia B.burdettii B.candolleana B.ericifolia B.dryandroides B.serrata B.prionotes B.baxteri B.lemmaniana B.caleyi D.ca¡linoides D.polycephela B.illicifolia B.baueri B.audax B.praemorsa B.meisneri B.occidentalis B.robur B.tricuspis B.pulchella B.grandis B.soland¡i B.coccinea D.formosa B.laevigata B.Iindelyana B.attenuata B.repens B.oligantha B.cuneata M.heterophylla Banksia RAPD phylogeny Figure 10.4 Strict consensus tree obtained from PAUP, tree length = 4079; consistency index = 0.929. Taxa represent two closely related genera Banksia and Dryandra, wirh Musgravea as the outgroup in the analysis. Tree generated from 791 RAPD bands produced with 15 selected primers. 165 B.ashbyi B.elderiana B.menzeisü B.blechnifolia B.candolleana B.bu¡denii B.integrifolia B.quercifolia B.laevigata B.lindelyana B.attcnuata B.meisneri B.occidentalis B.robur B.illicifolia D.carlinoides D.polycephela B.baueri B.audax B.praemorsa B.ericifolia B.dryandroides B.serrata B.prionotes B.baxteri B.ricuspis B.pulchella B.grandis B.soland¡i B.coccinea D.formosa B.lemmaniana B.caleyi B.repens B.oligantha B.cuneata M.heterophylla B anlcsia RAPD phyloqeny Discussion Both methods of analysis gave similar species pairs and groups, with closely related species within series, and closely related series generally grouping together, as in the classification of George (1981, 1988). It is interesting that Banksia coccinea grouped with species of. Dryandra in both analyses, indicating a close relationship. In addition Dryandra, in the presence of an outgroup, could not be clearly distinguished as a monophyleúc clade from Banksia using RAPDs. It is generally agreed that Banl Dryandra are sister taxa, with parallel developments in the two genera. Few morphological characters separate the two genera, so it could be suggested that Banksia andDryandramay be artificial genera. Pollen - pistil data with B. coccinea *ri!r1r:Ki species also show more compatibility than some Banl ¿-i++res+). Using RAPDs there were close relationships between species in series Banksia and Crocinøe, series Dryandroideae, Spicigerae, and Abíetinae, and series Bauerínae and Cyrtosrylis. The close relationship of series Crocínae and Banksi¿ has been shown previously using pollen - pistil compatibility and cladistic analysis (Sedgley et al., 1994, Thiele 1993). The series Spicígerae, Abietinae atdDryandroideae are grouped in section Oncosrylis, subgenus Banksia, confirming the close relationship of these series (George 1981). Subgenus Isosrylis formed a group with Musgravea, separate from the rest of Banksía showing more genetic distinctness than subgenus Banlcsia and genus Dryandra. RAPDs are not generally considered to be informative at the distantly-related species level. Concerns regarding RAPD generated phylogenies at higher levels include: (a) homology of bands showing the same rate of migration; (b) causes of variation in fragment mobility; and (c) origin of sequences in the genome. Knowing the identity of shared fragments is essential, since analysis depends on the number of shared bands. The vatidity of the assumption that fragments of the same size are homologous is still debated, and southern hybridisation of RAPD bands in other studies have shown varying levels of 166 B a nlcs i a RAPD phylogeny homology (Stammers et a1.,1995). Closely related species are expected to be more homologous than distantly related species, with more RAPD bands in close species corresponding to homologous sequences with conserved organisation. A consideration of change in fragment migration is important. The RAPD method relies on the relaxed conditions of primer annealing, and annealing is more critical at the 3' end than the 5' end, so base changes near the 3' end will significantly affect primer efficiency. It is likely that insertion/deletion events determine different fragment mobilities. The nature of the sequences detected by RAPDs are significant. Many bands contain repetitive domains (Stammers et a1.,1995), and it has been suggested that repetitive sequences are not suitable for phylogenetic analysis, because they can lead to concerted evolution. This may bias the phylogenetic tree due to the rapid spread of the variant repeated sequence. Although rhere are a number of potential problems associated with RAPD phylogeny analysis, there are also some clea¡ advantages. Many loci are examined at the same time, which minimises the effect of loci which are under selection pressure, giving divergence of the whole genome. RAPDs are usually considered to be appropriate for closely related groups, as shown by this study and studies on Loliun (Stammers et al., 1995), Medicago (Brummer et,a1.,1995), Lotus (Campos et al., L994), Brassica (Demeke er \,Jr\ikiq aI., 1992) and Allium ($lillçþ¿¡ al., 1993). IÓ7 Banl Chapter Eleven Application of non-coding chloroplast DNA sequences to B anksia (Proteaceae) PhYlogenY. Abstract Phylogenetic relationships within the genus Banksía L.f., (Proteaceae), and between Banksia and the related genus Dryandra, with Musgravea as the outgroup, were investigated using non-coding chloroplast DNA (cpDNA) sequences between ¡}re trnL (UAA) and. trnF (GAA) gene. The phylogeny obtained is targely in accordance with the traditional classification of subgenus Banksia, into two main sections Oncostylis and Banl two Dryandra species and the two sections of subgenus Banksia. Subgenus Isostylis and forms a polytomy with D . formosø, basal to subgenu s Banl genera with parallel development. Introduction 4he-gennsåønksia-I-f6with-over-?5taxa (George-1'98L'1'988)-isan-endemi€-AusE¿lian qrcrnber- sf-the-farn-ily*Ploteaeoae--R-elationships-amongst-thsProteaeeae-are-tentativq vttrffiøn1æiaataeed.in-the.subfemiþG'rer.illeoideae. Within'this'subfar'r.üy-thereare six tribes-ineluding-Bankis-ieae-and--Musgraveinæ;the-'latter-restieted-to-the-r*inforestsofi 'nodr-castern-Austa-lia..V/ithin tribe .Bankisieae there'are- two*genera:-Bønksúø-an* Ðryeúrq-each-oÊwhieh is.eonsidered to"be well-defîned; nat'trral-and r'nonqphyle*q dtheugh-$Èth-someparalel+erp@ 1óE Banksia ohvlosenv using cpDNA secuence $,ar €a+alogue&eÊspoeies-(Brow@ iee Wh fe-lirres-of-evoludonl i eie¿is¡ie+nelys,+ @+prev.ideôa.?oegnateh-r*ith+h - -suggest-a'robust-.ptrylogenetie-seheme{+was-norused-fora{orma}-elassi'fieatioR:-More- @) proposed a elassifieation baseé en eladistie analyses ef morpbologiealeha¡aote¡+*Aft hough+her. iffi oln+agreernent'q¡ith-+@ elas+ifi@Ì+åerc.a¡e+till+neeraintiesin"tho'positisns of.'somo - epeeieqeuehe+A= eeæi*e+ and++tWort-foF-some-nodes-is-tenuou*- Pa*¡eærlpri@ tåe+peeie"aæ+uped.rciailysi.mila¡-gÐuy¿+e "r€pese+het Is-a4úbmay-be-Aeser+o+ry¿@ rBam&súø.N@if as-*..s*bgerrlr s-øLBenks+'a;-with-the-eomment.t+rât-sepaffitÈ€Bneri€-s+â'ttls-ma:tåÈ appropriate--Ilis-iüerFet¿¡i.on-was-basedon-three-morphological-cha¡acte¡sJhe+hree specie€-in"subg€nss¿s@sryJrs-laok+he-prorninen¿in"voluoal-.braots-typioal'oÈDryandratbut 'h¿ve*flowolr"subtending-br¿ets typieal-oËB-ønksia;+heir{ollieles-arelåiek;rroody-an# +ementos€-as -ta.Ba*ksia-aad-they-havo.the-ovoiòin.floresoencc-Ðti.s-tJæiæLefÀankciø'- .Et6qry6¡r6¡..the¡eis*sti.l.l.-spoeulat-ion-on""+he-rola.tionship+-rn4tlún-*ubge*@ bet¡#e€ftsr¡bgenus -kasry/.r* *ubgenus.Bø*siø-and Dryatdra.@ må*er$rs^eå¿åsþte-res el+e-lù+ese*ela+ionsl+ip s" Chloroplast DNA (cpDNA) has been used extensively to infer plant phylogenies at different taxonomic levels. Direct sequencing of polymerase chain reaction (PCR) products is now becoming a rapidly expanding area of plant systematics (Clegg and I69 Banl Zurawski 1991). The rbcL gene encoding the large subunit of RUBISCO, has been widely sequenced from many plant taxa (Chase et aI., 1993). These phylogenies are particularly informative at the family level (Morgan and Soltis L993), and at higher taxonomic levels (Bousquet et aI., 1992). Phylogenetic relationships using rbcL sequences have also been used at lower t¿xonomic levels, indicating that it is useful at the generic level, however, in some cases the relationships remained unclear CXiang et al., 1993). The rbcL gene is considered to be too conservative to resolve relationships between closely related genera (Xiang et a1.,1993)- Anatysis of non-coding regions of cpDNA can potentially overcome the poor resolution of genes such as rbcL at lower taxonomic levels. These non-coding regions tend to evolve more rapidly than coding sequences, by accumulation of insertions/deletions at a rare equal to thar of nucleotide substitutions (Clegg and Zurawski 1991). cpDNA has been found to be extremely valuable for studying relationships between closely related species (Clegg et aI., 1991). Comparisons of the rates of rbcL and two non-coding regions of the cpDNA, the tnL (UAA) intron and the intergeneric spacer between the ynL (JAA) 3' exon and the rrnF (GAA) gene, have been studied in several genera and have shown that these regions evolve faster, providing greater resolution at the generic and intrageneric level (Taberlet et aI.,I99l, Ferris et aI., 1993, Gielly and Taberlet 1994). This study examines the phylogenetic use of non-coding cpDNA in Banksia, and between Banlcsia and the closely related gonus Dryandra, using the intergeneric spacer ben¡¡een thetrnL (UAA) andtrnF (GAA) gene. 170 Dhvloseny usinq cpDNA sequence Materials and methods Plant materínl Species included in the study are: Banksiø.' subgenus Banksia, section Banksía, B. serrata. (two specimens), B. medía, B. integríþlía, B. coccinea; section Oncosrylis, B. ericiþlia, B. spinulosa, B. sphaerocarpa; subgenus Isosrylis, B. cuneata, B. illiciþlia: Dryandra: D. formosa, D. carlinoides, D. polycephala; with Musgravea heterophylla as the outgroup. Seeds of Banksia and Dryandra were obtained from Nindethana Seed Service. Fresh leaf materi al of M . heterophylla was collected from Atherton, Queensland (QRS Arboretum No. 1372). DNA extraction DNA was exrracted from each seed using a modified method of \Meining and Langridge (1991), comprising: phenol/chloroform/isopropanol extraction for 5 min on ice, DNA precipitation for 1 min with ice cold isopropanol and sodium acetate, with DNA recovered by centrifugation at 12 000 rpm for 10 min. The pellet was washed twice with 70 Vo ethanol, dried and dissolved in 50 pL of TE buffer, with 1.0 pL of RNAase (R40: 0C 40 glmLRNAase A in TE), and stored at 4 0C short term (up to 1 month) or -20 long term (up to 1 year). DNA was extracted from fresh leaf material of. Musgravea heterophylia using the seedling leaf extraction method of Maguire et aI. (1994). The DNA was subjected to gel electrophoresis on a 1.67o agarose gel in TBE buffer (Sambrook et at.1989), and stained with ethidium bromide. DNA concentration was estimated by visual assessment of band intensities, compared to salmon sperm genomic DNA standards. The DNA content was adjusted to 10 ng pL -1. T7L Dhvloseny using cpDNA sequence DNA arnplification DNA amplification was performed in a MJ Research Thermal Cycler. The optimal program commencod with an initial step of 94 0C for 2 min, followed by 35 cycles of I min at 94oC,1 min 58 0C, 2 min 720C. Optimised reaction conditions were carried out in a25 pL total volume containing lXTaq buffer (Gibco-BRL), 3 mM MgCl2, 1 ¡tM of each primer, 200 pM of each dNTP (dGTP, dATP, dCTP, dTAP), I unit of Taq polymerase (Gibco-BRL), and 10 ng pL -1 template DNA. Each reaction mix was overlaid with PCR grade paraffin oil. DNA amplification fragments were separated by zEo agarose gel (Seakem, Promega) electrophoresis using TBE buffer (Sambrook er ø/. 19S9). A DNA size ma¡ker was used (pGEM, Promega) on each gel, and gels were stained with ethidium bromide. The intergeneric spacer between the ffnL (UAA) 3'exon and t r nF (GAA) gene was amplified using the primers e (GGTICAAGTCCCTCTATCCC) and f (ATTTGAACTGGTGACACGAG) (faberlet er al.,l99t) DNA sequencing Following amplification, excess primers and deoxynucleotide triphosphates were removed from samples using polyethylene glycol (PEG) precipitation in magnesium chloride (Nicoletti and, Condorelli 1993). Direct PCR sequencing of the purifred fragment was carried out using standard conditions in the DyeDeoxy Terminator Sequencing Kit of Applied Biosystems, then automatically sequenced on an Applied Biosystems Model 373A. Both directions of the non-coding region were sequenced, and a consensus sequence determined for each species. tlz Banksiq phylogeny using cpDNA sequence Dan øralysís Multipte alignments of the sequences of the non-coding region were performed manually. Phylogenetic relationships within Banksia, and between Banksia andDryandra, tsing Musgravea as the outgroup, were analysed via a parsimony approach using PAUP 3.1.1 (Swofford 1990). The phytogeny was assessed using the heuristic search method of PAUP (character optimisation ACCTRAN, MULPARS and TBR branch swapping options). To ensure that all islands of most parsimonious trees were found (Maddison 1991), the search was repeated 100 times with RANDOM addition and with a ma.: of 100 trees saved at each replication. Results Double stranded DNA amplification products were obtained for all species. The consensus sequence of the spacer region between ¡he trnL (UAA) 3' exon and the rrnF (GAA) gene is shown (Fig 11.1); the size of the aligned sequence for all species is 413 bp. There was no intraspecific variation between the ¡wo individuals of B. serrata . The single most parsimonious tree found using PAUP had a consistency index of 0.983, tree length 60, and a retention index of 0.909 (Fig 11.2). Generally, mostBanksia species fell into the two main sections of Banksia, subgen. Banksia, based on morphological characters: B. ericiþlia, B. spinulosd, and B- sphaerocarpa, of section Oncosrylis forming a clade with B. integriþIía. A second clade comprisedB. media and the two individuals of B. serrard, representing section Banksia. B. coccineø along with two Dryandra species D. carlirníde,r and D. polycephcla were unrcsolved at the polytomy with the two sections of subgenus Banksia. B. illíciþlia, Dryandra formosa and B. cuneata were basal to this polytomy with the related two members of Banksiø subgen. Isosrylis separated. 173 Banksia phylogeny using cpDNA sequence Figure 11.1 Complete nucleotide sequences of the spacer between ¡he trnL (JAA) and rrnF (GAA) gene tn Banksia, Dryandra and Musgravea (Length of alignment 413 bp). 1. B. serrata (individual t),2. B. serrata (individual 2),3. B. coccinea, 4. B. cuneata,5. B. eríciþIia, 6. B. il)icíþlia,7. B. integriþlía, 8. B. media,9. B. sphaerocarpa, 10. B. spinulosa,lI. D. carlinoides,12. D.formosa, L3. D. polycephala,14. M. heterophylla. 174 s 15 2s 35 45 55 ------+ +------+------+------+------+-_A-TAG-GTTCTTGGG- 1 O T_TCCCGÄ- -ATTCC_G-GC-TCATCCTCATT- -TATGTCAAT 60 2 O T-TCCCGACTATTCC-G-GCATCÀT- CTCÀTT__ACTÀG-GTTCTTGGGTCTATGTCÀÀT 60 a O T_TCC-GACTÀTTCCTGTGC-TCATCCTCÀTT-TA-TÀGAGTTCTTGGGTCTÃTGTCAAT 60 4 O T-TCCCGÀCTATTCCTGTGC_TCÀTCCTCÀTT-TÀ-TÀGÀGTTCT_GGGT-TÀTGTCAÀT 60 5 O TÄTCCCGÀCTATTCCTG-GCATCÀTCCTCÀTT_TÀCTÀGÀGTTCT-GGGTCTATGTCÄAT 60 6 O T-TCCCG- -TÄTTCC-GTGCATC_TCCTCÀTT-TA_TAG-GTTCTTGGGT_TÀTGTCAÀT 60 7 O TATCCCGÀCTÀTTCC_GTGCÀTCÀTCCTCATT-TÀ_TÀGAGTTCTTGGGTCTATGTCAAT 60 I O --- CCC-ÀCTÀTTCC-G-G-_TC,ATCCTCATTTT.ACCA-ÄGTTCT-GGGTCTATGTC-ÀT 60 9 O T-TCCCGACTATTC_-G-GCÀTCATCCTCATT-T-ATAG-GTTCT_GGGT_TATGTCÀÀT 60 10 O TATCCCGACTATTCCTGTGCÀTCÀTCCTCATTCTACTAGAGTTCT-GGGTCTÀTGTCÄÀT 60 l_1 O ---_ CCGÄ--ATTCC-G-GC-T-ÀTCCTCATT--,4-TÀGÀ_TT_TTGGGT-TATGTCAAT 60 _TGGGT-TA-GTCÀÀT 60 1_2 O -- - -- CGÀ-T-T- CC_G-G-_T_ -TC-T_ ATT-TA_T- -_G-T- 13 O T-TCC_GA_TATTCC_G-GCAT_ÀTCCT_A-GA_À-TAG-GTT--TGGGT-TATGTCAAT 60 I4 O G-TCCOGA-_-TT_ C_G-G- _TC_TCCTC_TT-TA- -À-AGG-T-TTGGG-TATG-C-AT 60 65 75 85 95 l-05 115 ----+------+------+------+------+ +----- 1_20 1_ 6 O TÄAAGGÀTTÀAAAAAACATTÀCÀAAGTCTTA- CCCAGGCCCCGGAAATTCTTGGATCTTC 2 6 O TÀAÀGGÀTTÀAAAAAACÀTTÀCAAAGTCTT,A_CCCAGGCCCCGGÄÀATTCTTGGATCTTC r20 3 6 O TAÀAGGACTÀAAAÀAÀCATTÀCAÀÀGTCTTÀ_CCCAGGCCCCGGÀAATT-TTGGAT-TTC L20 4 6 O TAAAGGACTAAÀÄAAACATTACA,AÀGT-TTÀTCC-AGGCCCCGGÀAÀTT-TTGGAT-TTC 1_20 5 6 O TÀAAGGACTÄAAÀAAACÀTTÀCAAAGTCTTA-CCCAGGCCCCGGAAATT-TTGG.ATCTTC r20 6 6 O TÀAÀGGÀCTÀAÀÀÄAACATTACAAAGTCTTÀTCC_AGGCCCCGGAÄATT-TTGGAT-TTC L20 7 6 O TAAÀGGÀCTÀÀÀÀAAÀCATTÄCÀAÀGTCTTÀ- CCCÀGGCCCCGGAAÀTTCTTGGATCTTC 1,20 I 6 O TÀÀÀGGÀTTAAÀAAAÄCATTÄCAÀAGTCTTÀ-CCC__GCCCCGGÀÀATTCTTGGATCTTC L20 9 6 O TAA-GGACTAAÀAAÀ.ACATTACAÀAGT_TTA_ CCCÀGGCCCCGGAÀCTT-TTGGAT_TTC 1-20 10 6 O TÀÀAGGACTAÀÀAAÀÀCÀTTÀCÀÀAGTCTTÀ_ CCCAGGCCCCGGÀÀÀTTCTTGGATCTTC 1"20 11 6 O TAAÀGGACTAAAÀÄÀÀCÀTTÀCAÀAGTCTTA-CCCÀGGCCCCGAAÀÀTT-TTGGÀTCTTC L20 L2 6 O T-Ä-GGÄ_TAÀè.4ÄÀÄCATTÀCAÄA-T-TTA_CCCÀGGCCCCGGAÀ-TT_TTGGÀTCTTC L20 t_3 6 O TÀÀAGGACTAÀAAÀAACATTÀCAAÀGTCTTÄCCCCÄGGCCCCCGÀAATTCTTGGATCTTC r20 L4 6 O TÀÀAGGGCTÀÀÄAAAACATTÀCAÀÄGT-TTA-CCCAGÀCCCCGGGÀÀTT-TTGGGTTTTC 1_20 L25 135 145 155 165 1-75 ----+------+------+------+------+ +----- 1 1_20 AAÀ.ÀAGAAGACTTTGTAAGTTT CATTAÀGÀTTAAGÀGTAATTATATGGATGGTÀÀATGAT 180 2 r20 .AAÀÀÀGAÀGÀ CTTTGTÀÀGTTT CATTÀAGÀTTÄAG.AGT.AÀTTATATGG.ATGGTÄAATGAT 180 3 120 AAAAÄG AAGÀ CTTTGTÀAGTTT CÀTTAÀGATTÀÀGAGTÀÀTT.ATÀTGGATGGTÀÀATÀÀT 180 4 L20 ÀÀAÀÀ- ÀÀGÀCTTTGTÀÀGTTT CÀTTÀA- ÀTTA_ _ À_TAÀTTATÀTGGÄTGGTÀÀATGAT l_80 5 r20 ÀAAÀAAAAGACTTTGTAÀGTTT CATTAAGATTÀAAÀGTAATTÀTÀTGGATGGTAAÀTGÀT 180 6 1_20 AAAAAGAAGÀCTTTGTAÄGTTTCATTÀAGATTÀAG,AGTAÀTTATATGGATGGTAÀATGÀT 180 7 t20 ÀÀAÀAG ÀÀGÀ CTTTGTÀAGTTT CÀTTAAG ÀTTÀAGAGTÀATTATATGGÀTGGTÀAÀTGÀT 180 o L20 ÀÀAÄÀ- AÀ- ACTTTGTÀÄGTTTCÄTT- -GÀTTÀÀ- AGTAATTAT-TGGATGGT- -ATGAT 180 _ _,ATGGTAAÀTGAT 9 L20 AÀÀÀÀGÀÀGACTTTGTAÀGTTTCATTAÀGATTAÀGÄGTAATTÄ_ - - 180 10 1_20 ÀÀAÀAGÄAGA CTTTGTAAGTTT CÀTTÀ,AGATTAAAÀGTÄATTÀTATGG.ATGGTAAATGÀT 180 11 t20 AÀÀÀAGÄÀGÄCTTTGTAÀGTTTCÀTTAÀGÀTTAAGÀGTÀÀTTATATGGÀTGGTÀÀÀTGAT 180 1_2 L20 AAÀÀÀG ÀÀG,A.CTTTGTÀÀGTTT CÀTTÀAG ÀTTAÀGÀGTAATTATATGGÀTGGTÀÀÀTGAT 180 1_3 L20 AÀÀAÀG ÀÀG ACTTTGTAÀGTTT CATTAÀG ATTAÀGAGTÄÀTTATÀTGGATGGTÀAATGAT 1-80 t4 L20 CAAÀ-GG- _ A- - TTGTAÀGTTT- ÀTTÀÀGGTTÄÀGGGT-ÀT_,4TÀTGG-TGGTÀAATGÀT 180 18 5 19 5 205 21-5 225 235 ----+------+------+------+------+ +----- t_ 180 TTAGATTCAAATGGGAATTCCTTG CTCATÀGATGTTCATTTGTACGTATATCTTAATATA 240 2 180 TTÀGATT CÀÀÀTGGG.AÀTTC CTTG CTCÄTAGATGTT CÀTTTGTÀCGTÀTÀT CTTAATAT.A 240 3 t_80 TTAGATT CÀÀÀTGGGÀÄTT CCTTG CT C.ATAGATGTT CÀTTTGT,ACGTATAT CTTÀATATÀ 240 4 t_80 TTé.- ÀTTCÀAATGGGAATTCCTTG CTC.ATÀAÄTGTTCÀTTTGTA-GTÀTAT-TTÀÀTATA 240 5 180 TCAGATTCÀAATGGGÀATTCCTTG CTCATÀGATGTTCATTTGTACGTATATCTTAATATÀ 240 6 180 TTÄGÀTTCAAATGGGAÀTTCCTTG CTCATÄAATGTT CATTTGTÀCGTÄTAT-TTAATATÀ 240 7 180 TCAGATTCAAÀTGGGAATTCCTTGCTCA,TÀGÀTGTTCATTTGTè.CGTATÄTCTTÀATÀTA 240 B t_80 -TÀGATTC- --TGGGÀè.TTCCTTGCTCÀTAGÀTGTGCÀTTTGTACGT.ATÀTCTTAA^TATA 240 9 180 TTAG,\TTCAAÀTGGGAATTCCTTG CTCATAGATGTT CÀTTTG- A- - -ATÀT=TTAAT-TÀ 240 10 180 TCAGATT CAÀÀTGGGÀATTCCTTG CTCÀTAGÀTGTT CATTTGTÄCGTATATCTTAAT.ATA 240 11 t_80 TTAGÀTT CÀAATGGGÀÀTTCCTTG CTCATÀGÀTGTTCÀTTTGTÀCGTATÀTCTTÄATÀTÀ 240 L2 t B0 TT.\G ÀTT CAÀÀTGGGÀ.ATT CCTTG CT CÀTÀGATGTT CATTTGTA- GTA,TÀ.T -TTÀÀTAT.A 240 13 180 TTÀGATTCÀAÀTGGGAÀTTCCTTG CTCATÀGÀ.TGTTCATTTGTACGTÄTA,TCTTÀÀTATÀ 240 1-4 l_80 T- AGGTT- ÄAÄTGGGGÀTTCC- TG_ T_ ATÀGA-GTT-ATTTGTÀ-G- -TÀT-TTÀGTATÀ 240 245 255 265 275 285 295 +------+------+------+------+------+ 300 1_ 240 TCÄCATÄTCACÀAGACTTGTGGTAAGA_GAGAAÄGATTTCTGCTCGGÄTCCATTTGT-GA 2 240 TCACATÀTCÀCAAGÀCTTGTGGTAÄG,A-GAGÀÄ.AGÄTTTCTGCTCGGÀTCCÄTTTG- -GA 300 a 240 TCÀCÀTÄTCACAAGÀCTTG,AGGTÄAGA_GÀGAAÀGATTT- TGCTCGGÀTCCATTTGT-GÀ 300 4 240 T CÀCÀTATCÄCAAGÀ-TTGTGGTÀ.AGA-GAGÀAAGATTT_ TG-T_ -GÀTCCÀTTTGT-G,q 300 5 240 TCACATÀTCACAAGÀCTTGTGGTÄÀG.A-GAGAAÀGATTTCCGCTCGGÀTCCATTTGT-GA 300 6 240 TCÀCÀTATCÀCÀÀGACTTGTGGTÀÀG.A-GÄGAAÀGÄTTT-TGCTCGGÀTCCATTTGT-GA 300 240 TCACATATCACAÄGACTTGTGGTAAGA_GÀGAAÀG.ATTTCCGCTCGGATCCATTTGT_GÀ 300 B 240 T CÄCÀT- TTÀCÀÀGACTTGTGGTAÀGÀ_G.AGAÀÀGÀTTTCTGCTCGGATCCATTTGT-GÀ, 300 9 240 TCACATATCACAÄGACTTGTGGTAÀGÄ-GAGÀÀÀGé.TTTCCG CTCGGÀTCCÀTTTGT-GÄ 300 10 240 TCACATATCACÀAGACTTGTGGTAÀGÀ-GÀGÀÄAGÀTTTCCG CTCGGATCCATTTGT-GA 300 11 244 TCACÀTÄTCACAAGÀCTTGTGGT.AAGA-GAGAAÀ- ATTT- TGCTCGGATCCATTTGT-GA 300 L2 240 TCÀCÀTÀTCÂCAÀGACTTGTGGTÀÀGÀ.-GAGAAÀ- ATT-T-GCT- -G.ATCCATTTGT-GÀ 300 _T-GA 13 240 TCACÀTÀTCÀCÄÄGÀCTTGTGGTAÀGA- - A- ÀAA- ATTTCTGCTC-GATCCATT- 300 1_4 240 TCÀ- ÀT_ TCACÀAGÀATTGTGGTÀ.AGA-À-GAAÀ- ÀTTT- TG-T- - T- T- CCCTTGT-GA 300 305 315 325 335 345 355 ----+------+------+------+------+ +- ---- 1 300 AÀGAÀAÀGAÄAÀAGÄÄTAGTÄGAGTGÄATGAGAÀACATÀACTÀÄATTTGAGÀAGGAGAAC 360 2 300 ÀAGÀÀAÀGÀÀÀÀÀGAÄTÀGTÄ_ _ _TGAÀTGÀ_ AAA CÄTÄÀCTÄÄ- TTTGÀ-AÄGGA-Ä,AC 360 3 300 ÀÀGAÀÀÀGÀÀÀÀAGÄÀTAGT.A- À_ TGÀÀTGÀGÀÀACATÀÀCTÀAATTTGAGÀÀGG.A-AÀC 360 _ 4 300 - _GGAAÀGÀÀÄA-GAAÀAÀTT- ÀGTGAATGÄ- ÀÀÃ_ -TAA-TAÀATTTG,À'-AAGGÄ- AC 360 5 300 ÀAGÄAAAGAÀAAAGÄÄTA- TA_ AGTGÀÃTGÀGAÄACÀTÄÀCTAÀATTTGAGAÀGGÄ_ ÀÀC 360 6 300 ÄÀGÀÀÀÄGÀAAAÀGAÀTÀ_TA_ _ -TGÄÀTGA-ÀAÀ CATÀÀCT.AÀATTTGAGÀAGGÀ_AAC 360 300 A AGÄÀAÀGÀÄAAAGÀATÀGT,AGÀGTGAATGAGÄAA CÀT.AÀ CT.AAATTTGAGAAGGAGAÀ C 360 I 300 ÀÄGÀÀÀÀGÀ,AÀAÀGÄÀTAGTA_ ÀGTGAÄTG.AGA_ À CATÀACTÀÀ,ÀTTTGAGA,AGGÀGÀÀ C 360 9 300 AAGAÀÄÀGÀAAAÀGAATAGTA- AGTGAÄTGÀGAÀACÀTÀÀCTÀÀATTTGÀGÀAGGÀGAAC 360 10 300 AÀGÀÀÀA- _ _ _ - ÀGÀATÀGTÀGÀGTGAATGÀGÀÀACÀTAÄCTÀÀÀTTTGAGAAGGÀGAAC 360 11 300 AÀGÀ,AAÀGAÀÀÀÀGÄATA- TÄ- Ä- TG,AATGA_ ÀAACÀT,A_ CTAA- TTTGÀ_ AÀGG_ - ÃÀC 360 L2 300 Àè.GÀAAAGÀÀ,AÄAGAÄTÀ-TA- A-TGAÀTGA- À.ÄÄCATA_ -TÀÀ- TTTGA-ÄAGG,A_AAC 360 1a 300 AA-AAÄÀGÀÀAAAGÀAT- -TA-_ -TGAÀT_ - _ÀÄAC-TÀ- CTÀÀ-TTT-A-ÀÄGGA-ÀAC 360 14 300 AG- AAÀ-GÄÀAAÀAÀAÀ- ATÀTTGTGÀATGAGAAÀTATAÀ-TAÀ_TTTGG-ÀGGGA- - _ - 360 365 375 3B5 395 405 ----+------+------+------+------+ 41_3 l_ 360 GATGACTAÄÀTTGGÀATCGCTGÄCGAÀÃAÀAAA_ - -TTÀGGGÀATAA_ CCGGG _GG 2 360 GÀ-GACTÀÀ_TTGGÄ-- C- CT_ÀC-ÀAAÄAAAAA- -TT_GGGAA- _Ä- C_ 41_3 3 360 GÀ-GÀCTÄÀ-T-GG,AATCG CTGÀ_ - ÄAÀAAÀAAA_ _TT-GGGAÀ- - 4]-3 _- _-GG 4 360 GÀTGÄCT- -ATTGG.\A_ C_-TGÀC-ÀAAÀÀÀAÀÀ- _TT---_- -_ - - 41_3 _TTAGGG- _ 5 360 GÀTGÀCT,AAATTGGAATCGCTGACGÀÀÀÀÀÀÀAG- - - 41_3 6 360 GÀTGÀCTÀA_TTGGÀ-TC-CTGÀC-,AÀÀÄÄÀÄÀ_-- AÀ_TT__- 41_3 _GG 7 360 GÀTGACTÀÀATTGGAATCG CTGACGÀÀAÀAÀÀAA- - TTÀGGGAÄTAÀ- C- 4L3 _ _ _ B 360 - _TGACTÀAÀTTGGÀATCG C-G.AC_ AAÀÀAÀA- - 41-3 9 360 G-TGÀCT- - -- -GGAATCGCTGACGÀÀÄÀÀ^AAAAA_TT-GGGG_-- -ACCGGG 4L3 10 360 GATGACTAAATTGGAATCGC-GACGAÀÄ.4.4ÀAÀÀ_GTTAGGGAATÄÀACCGGG 41"3 1l- 360 _À__À_ - ÀÀ-TTGGÀ- - C- C- - _ C- AÀÀÀAÀAÄA-GT- _- - _ _- - 41"3 L2 360 _À__ CC- AA-TTGGAA- - - CT-- --AAÀAÀAAAÀÄGTTAGGGAATÀAACCGGG 4t3 l_3 360 -À--A-TÀA-TTGGA--C--__-C_AAÄAÄÀAA--_ À__--__- 4L3 T4 360 G- TGA- TÀÀ_TTGG.AATCG CTGACGÀÀAAÄÀAÄAAATT-GGG,AÀTÀA- CC- - - 4L3 Banksia phylogeny using cpDNA sequence Figure 11.2 Phylogeneric relationships wittrin Banl 175 B.serrata B.media B.serrata B.coccinea D.carlinoides D.polycephala B.ericifoIa B.spinulosa B.integrifolia B.sphaerocarpa B.illicifolia D.formosa B.cuneata M.heterophylla Banlcsia phylogeny using cpDNA sequence Discussion The trnLintron and the spacer between tnL and trnF are valuable tools for preliminary studies in closely related species goups that have not been investigated previously. These primers are universal and amplify DNA from a wide range of species, and due to the small size of this region, no internal primers a¡e needed for sequencing, so information is easily obtained. The phylogeny obtained using cpDNA sequence data show broad species grouping into the two sections of subgenus Banksia, section Banksia and section Oncostylis. B. coccinea,whose relationships are still uncertain, is distinct from either section grouping instead in a polytomy with two species of. Dryandra at the node with the two sections of subgenus Banksia, thus suggesting that a thfud section containing B. coccineø may be more appropriate. This separation has already been proposed rece¡tly, based on Chaøler I morpho1ogica]charactersandpistil-pollencompatibilitydata(@.The node between the rwo sections of Banlcsia with B. coccínea and the two Dryandra species is unresolved, and the relationship of B. coccinea to the Dryandra species is unclear. The non-coding region between trnL andffnF appars to be too conservative to clearly resolve these relationships, and a faster evolving region might be more appropriate. Subgenus Isosrylis and D. formosa were basal to the remainder of Banksía and Dryandra. George (1981) indicated that subgenus lsosry/rs may be more closely related to Dryandra than Banksiø, but placed itin Banksi¿ with a note that a new genus may be appropriate. Interestingly, Dryandra ín the presence of Banksía, did not separate as a monophyletic clade based on its cpDNA sequence data. These data suggest that the currently accepted view of two sister genera, Banksía and Dryandra, with parallel morphological development, may be inappropriate. Based on cpDNA sequence data, it can be suggested that Banksia and Dryandra are possibly artificial genera. Similarly, molecular data have suggested that widely accepted views of separate genera may be 1'ló Banksia ohvlosenv usine coDNA se{uence artificial in otherplant groups such as Eucalyprzs (Sale et aI., L996, Ladiges et al., 1995). The results of this study do not represent a def,rnitive study of the relationships between the genera Banksia and Dryandra, but it raises important questions about the cu:rently accepted views of the classifrcation. It is not easy, for many reasons, to establish the correct choice of a region of the chloroplast genome for resolving phylogenies. råcl has been shown to be valuable at the family level (Morgan and Soltis 1993), and at higher levels (Bousquet et a1.,1992). Phytogenetic relationships using råcl sequences have also been used at lower taxonomic levels, however in some cases the relationships remained unclear (Xiang et a\.,1993). Therefore, the rbcL gene is sometimes too conservative to clarify relationships between closely related genera. For this reason, non-coding regions of the cpDNA such as the trnL intron, the spacer between the trnL and trrF gene, and the spacer between rbcL and atpB (Taberlet et al., I99I, Goldenberg et aI., 1993) have been advocated as being more appropriate for working at lower taxonomic levels. For large scale phylogenies of a genus or genera, a preliminary study, such as the one presented here, is advisable. This is because sequsnce divergence of these regions at the intrageneric level can vary greatly. In some cases the evolution of these regions is scarcely faster than the rbcL gene (Gielly and Taberletlgg4), and few comparisons of the evolutionary rates of both cpDNA and nuclear ribosomal DNA (nrDNA) have been conducted. Gielly et aI., (1996) compared the rates of the non-coding region of nuclear internal transcribed spacer regions (ITS) to the non-coding regions of cpDNA trnl (UAA) intron sequences in the genus Gentiana. Comparisons of the evolutionary rates found that sequence divergence in the ITS regions were higher than for the trnL introns. However, the cpDNA intron and the ITS of the nrDNA gave largely concordant phylogenetic trees. At the intrageneric level in Gentian¿, ITS sequences appeared to be more appropriate in the assessment of plant phylogeny, but the cpDNA rrnl intron was preferable at the intergeneric level. In our study, we similarly found trnL to be L]1 Banksia ohvlosenv usine cpDNA seouence consenative at the intrageneric level in Banksia and Dryandra.This again suggests the two may be artifrcial genera, challenging the accepted views of two monophyletic, sister taxa, with considerable parallel morphological development. This work shows that further investigation both into the taxonomic relationships within Banksía and between B attl 178 General discussion Chapter Twelve General discussion Banksia breeding Research is essential to understanding the biology of banksias and to develop strict selection criteria for crop improvement. A breeding program has been established for Banksia (Sedgley et al.l99t) which focuses on the species B. coccinea, B. menziesii, B. hookeriarø, and B. prionotes. Breeding proglams on other members of the Proteaceae have been established in South Africa and America, and have used interspecific hybridisation as the basis for cultivar development (Parvin 1981, Brits 1985). The Banksía program (Sedgley et aI.I99I,1994) also aims to use interspecific hybridisation within the genus to produce novel cultiva¡s and combine cha¡acters from other species. Hybridisation methods have been developed for Banksia based on knowledge ofthe breeding biology ofthe genus fuss and Sedgley 1991). In addition, research into the structure of the pollen presenter and the stigmatic gloove has shown that there is more than one type of pollen presenter, and the location of the stigmatic groove is different depending on the species (Sedgley er ø/. Igg3). This is imporrant in interspeciflrc hybridisation to make sure the target is clearly identified for deposition of interspecifrc pollen into the stigmatic gloove. This thesis describes an extensive interspecific pollination program based on B. coccinea (chapter 4). To focus breeding efforts, species relationships need to be known in order to increase the probable success of hybridisation. It is generally assumed that closely related species will hybridise, and distantly related species will not. Interestingly, B. coccineais unique in the genus with distinct morphological characters and unclear relationships to other species. Interspecific hybridisation was conducted with a broad ra¡ge of species with B. coccinea as the male parent in order to determine its relationships, and secondly, to 179 General discussion identify species combinations that may be successful for hybridisation. A number of important conclusions arose from this work. Hybridisation success generally depends on species relationships, and only closely related species have compatible pollinations, as assessed by fluorescence microscopy. B. coccinea rwas most compatible with species of section Oncostylis, with some compatibility with species of section Banksia and subgenus lsosrylis. In addition, pollen tube growth was controlled in the pollen presenter and upper style regions, which has been demonstrated previously in inna and interspecific crosses (Sedgley et a1.,1994, Fuss and Sedgley I99l). Pollen germination is also controlled on the stigma, perhaps a more severe display of an interspecific incompatibility mechanism. Interspecific crosses with B. coccinea displayed pollen tube abnormatities not previously described inBanlcsia. These tpes of abnormalities have also been noted in the interspecific crosses of other genera @llis et al., L99L,Fntz and Hanneman 1989). Unilateral incompatibitity was not observed in crosses with B. coccinea. Unilateral cross incompatibility is commonly seen in other crops where a self incompatible female rejects the pollen of a self compatible male. It was suggested that unilateral incompatibility was not present in Banl of Banksia are largely outcrossing and thought to be self incompatible (Scott 1980, Carthew et a1.,1988, Goldingay et a1.,1991). Species combinations identified as potentially compatible were assessed for seed set. B. coccinea was crossed with species from section Banlcsia, section Oncosrylis and subgenus IsosryIis. Crosses failed to produce follicles and seed with species of section Banksia and subgenus Isosrylis. Crosses with species of section Oncosrylis formed follicles with aborted seed, and follicles with seed. Seed were tested for viability and seedling growth. The seed had low viability, with only two seedlings produced, one of which died. Based on early seedling morphology the remaining seedling was indistinguishable from the female parent. As it reaches maturity, it may display more intermediate chatacters, although the balance of probabitity suggests that it arose from self pollination. The cross that initiated 1E0 General discussion follicle set, but not seod, indicates that post fertilisation selection occurs in Banl Pollen storage and viability testing methods are important adjuncts to a plant breeding program. Stored pollen can be used to hybridise species that do not flower at the same time, or are geogaphically isolated, increasing flexibility of the breeding program. Pollen storage and viability testing methods were devetoped f.or B. menziesü, which are also applicable to other species (chapter 3). Two methods of viability testing were investigated, sraining with FDA and invitro germination. Staining methods were found to be unreliable and in vitro germination was preferable, with optimal conditions for in vitro germination developed f.or B. menziesíi. Pollen storage above zero degrees was found to be unsuccessful, with pollen retaining good viability at temperatures below zero. Pollen, after drying over silica gel, can be stored up to six months in an ordinary'fridge or freezer to suit breeding purposes. Genetic diversiry The continued interest in Australian species, particularly those from family Proteaceae, for cut flowers provides a challenge to plant breeders to supply new cultivars. Therefore, continued selection and cultivation of new variants and species is essential to expand curent markets and maintain high quality products. Conservation of natural populations of indigenous plants will ensure there a¡e sufficient resources to tap for further crop development and improvement in the future. 181 General discussion Traditionally, studies of genetic diversity have used allozyme markers. Allozymes can provide informative markers for levels of genetic variability within a plant species, but there are widely recognised limitations. Many species of the family Proteaceae have low levels of detected allozyme diversity. This is because the detection of variability is limited to protein coding loci, which may underestimate levels of genetic diversity (Clegg 1989), and may not reprosent the entire genome (Schaat et al. L99I). Allozymes ¿ìre also tissue specific and are influenced by environmental factors. Given these limitations many workers have moved to using DNA ma¡kers such as RAPDs. RAPDs overcome the limitations of allozymes and are simpler and easier to use than other DNA markers such as restriction fragment length polymorphisms (RELPs). RAPDs produce potentially unlimited numbers of markers and have been shown to be useful for a variety of applicaúons including population genetic studies on a number of genera (Haff et al. lgg3,Chalmers et a|.1992, Russell et al.1993). Genetic diversity has not been widely studied in Banksía. Allozyme methods have estimated. high levels of genetic diversiry in Banl for plants (Scon 1981, Carthew et al.,19SS). Genetic diversity for a number of species of Banksia using RAPD ma¡kers were investigated, in particular B. cuneata, a rare and endangered species. High levels of genetic diversity were detected in 33 species of Banksia, and three species of Dryandra using RAPDs (chapter 8), and levels detected using RAPDs were comparable to levels of within species diversity detected by allozymes. RAPDs also produced large numbers of polymorphic ma¡kers, compared to the limited number available for allozyme techniques. Genetic diversity of all species tested including threatened, or rare species, was high indicating that none have reached dangerously low levels. More research is needed however, to target fragmented species across all remaining populations. B. cuneata is declared a rare species under the WA Wildlife Conservation Act, and is under threat of extinction due to small population sizes and fragmented distribution over L62 General discussion private and public land. Genetic studies including all known remaining populations have not previously been conducted. Levels and patterns of genetic diversity were invesúgated using RAPDs (chapter 9), which is important for the future conservation and management of B. cuneata. Genetic diversity of B. cunear¿ estimated using RAPDs was quite high, considering it is a geographically restricted rare and endangered species. The 10 known populations of B. cuneata showed differences in the extent of disturbance and levels of genetic diversity. Variation in genetic diversity estimates between populations is nor uncommon in species with mixed mating systems (Schoen L982). Pollinator availability and activity are likely to have an influence on genetic diversity between populations, and this particularly applies to bird pollinated species such as B. cuneata- Population structure such as plant density, and level of disturbance may also influence pollinator behaviour, and this is reflected by larger populations with lower levels of disturbance, having high levels of genetic diversity. In contrast, small populations on road verges with high levels of disturbance, display lower levels of genetic diversity. The estimates of genetic diversity within populations were in agreement with allozyme methds conducted on 6 B. cuneata populations (Coates and Sokolowski 1992). Dissected population structure and geographic restriction of B. cuneatals likely to influence the patterns of genetic variation within and between populations (Hamrick and Godt 1989). As B. cuneata is outcrossing and bird pollinated, most va¡iation would be expected to be within populations and little between populations. Analysis of molecular variance (AMOVA) was used to pafiition RAPD variation within and between populations of B. cuneata- The results clearly demonstrated that nearly all variation was attributable to individuals within populations, with no significant genetic differentiation between populations. The lack of population differentiation in B. cuneata using RAPDs contrasts to a previous sfudy by Coates and Sokolowski (1992), using six of the ten populations of B. cttneata, and enzyme electrophoresis. Their study using 6 polymorphic loci, found significant differentiation of populations into two groups. It was further 183 General discussion suggested that differentiation may be due to a salt river system acting as an ecological barrier to pollinator movement. Direct comparisons of RAPD and allozyme diversity are few. Peakall et aI., (1995) compared RAPD markers to allozyme markers in buffalograss. Using AMOVA analysis, they found that both markers gave similar qualitative patterns, large regional differences and significant population differentiation within regions for four populations. RAPDs however, detected more variation than allozymes. Dawson et aI., (1993) found patterns of RAPD variation in Hordeum spontaneum comparable to previously reported patterns using allozymes. Liu and Fumier (1993) using two species of Populus compared RAPD, RFLP and allozyme markers. They found that RFLP and allozyme markers gave comparable pattems within and between species, while RAPDs revealed more va¡iation between individuals within a species than allozyme and RFLP markers, and produced reliable discriminarion among clones. The findings reported here with Banksia are consistent with reports that RAPDs detect more variation than allozymes (Peakall et aI., 1995, Liu and Furnier 1993, RusseLl et al., 1993, Dawson et aI., 1993)- Furthermore, the study using RAPDs on B. cuneør¿ shows nearly all variation is within populations rather than between. RusseII et al., (1993) also reported that RAPD diversity in Theobroma cacao was higher within provenances than between, a pattern typical for outcrossing woody plants (Hamrick and Godt 1989). Although allozymes a¡e valuable for many types of studies, it is clear that DNA markers such as RAPDs will be widely used on plant population studies in the future, and will remain a valuable genetic tool. Species rebrtonships For focused and efflcient breeding efforts, species relationships need to be known in order to increase the probable success of hybridisation. Species relationships were investigated using three approaches, interspecific hybridisation (chapter 4), RAPDs (chapter 10) andDNA sequencing (chapter 11). 184 General discussion Interspecific hybridisation found that.B. coccinea was more closely related to species in section Oncosrylis, than species in section Banksia, where B. coccineø is currently placed. Due to the distinct morphology of B. coccinea, and the lack of compatibility wittt section Banksia, it was proposed to move B. coccineø into a new section Coccinea, containing only B. coccinea. Furthermore, it was suggested that the new section Coccíneais the sister sectidn to Oncostylis, given the close compatibility of .8. coccinea to this section, shown by fluorescence microscopy and seed set data. The close relationship of. B . coccinea to section OncosryIis has been previously suggested based on morphological characters and cladistic analysis (George 1981, Thiele 1993). RAPDs wero assessed for phylogenetic signal at various levels within Banlcsia, and also between Banksía and the closely related genus Dryandra- RAPDs have been recently used in other genera, and have produced phylogenies which are in agreement with existing classifications based on morphology, cytology, and allozyme electrophoresis methods (Demeke et a1.,1992, Wilikie et a1.,1993). Debate continues as to which is the best method of analysis for RAPD data in phylogeny assessment, so two common methods of analysis were used on Banksia data. Both clustor analysis and phylogenetic inference using parsimony gave similar patterns in Banlcsía. Closely related species within series formed groups, and closely related species between series formed groups. These groups generally agreed with the current classification of Banl 1981, 1988). RAPDs were uninformative at higher levels between distantly related species. Interestingly, B. coccinea formed a group with species of. Dryandra, indicating a close relationship to this genus. Interspecific hybridisation also showed some compatibility with Dryandra, with one c¡oss showing more compatibility than some Banl than Banl be appropriate (George 1981). Few morphological characters separate subgenus Isosrylis from Banlcsiø andDryandra.Using RAPDs subgenus IsosryIís grouped w\th Musgravea, included as the out-group, indicating more genetic distincnress than subgenusB anlcsia and Ið) General discussion species of Dryandra. The three species of Dryandra did not sepamte as a group, and were indistinguishable from B anksia. DNA sequencing of cpDNA (chapter 11) also showed the same trends observed with interspecif,rc hybridisation and RAPD data. Using DNA sequence data, genus Dryandra was indistinguishable from Banksia.B. coccineø grouped with species of Dryandra, between the two sections of subgenus Banksia, section Banksia and OncosryIis. Resolution at the node between the sections with B. coccinea and Dryandra was poor using the region studied. A faster evolving region may be more appropriate to resolve the relationship between B. coccinea andDryandra. DNA sequence data support the placement of B. coccineaínto its own section, however its relationship to Dryandra is still unclear. DNA sequence data also support subgenus Isosrylis as distinct to subgenus Baricsia, with the ¡wo related species of subgenus lsosrylr"s and Dryandraformos¿, basal to subgenu s Banksia. Both molecular and sexual compatibility data suggest that Banl and, Dryandra may be artificial genera, challenging the accepted view of two natural monophyletic, sister genera, with parallel morphological development. In conclusion, this research contributes information important for conservation and exploitation of natural genetic resources for plant breeding. The relationship of. B. coccinea to other species of Banksia is more clearly resolved, however this work raises important questions about the currently accepted view of Banlcsia andDryandra. 186 References cited References cited Abo-elwafa, 4., Murai, K. and Shimada, T. (1995). Intra and interspecifrc va¡iations in Lers rcveafed by RAPD ma¡kers. Theoretícal and Applied Genetics 90,335-340. Ai, Y., Singh, 4., Coleman, C.E., Loerger, T.R., Kheyr-Pour, A. and Kao, T.H. (1990). Self-incompatability inPetunia inflata: isolation and cha¡acterisation of cDNAs encoding three S-allele-associated proteins. Sexual Plant Reproduction 3, 130-138. Akihama, T. and Omura, M. (1986). Preservation of fruit tree pollen. Biotechnology in Agriculture and Forestry. 1, 101-1 12. Amstrong, J., Gibbs, A., Peakall, R. and Weiller, G.F. (1994). RAPDistance: random amplified polymorphic DNA analysis. Computer program distributed by the Australian National University. htç: //life.anu.edu.auffolecular/Sofnvare/raPd.htnl. Anderson, M.4., Bacic, A., Clarke, 4.E., Ha¡ris, P.J., Hoggart, R.M., Man, S-L-, Rae, A. and Williams, E.G. (1984). Biochemistry of the pollen-pístíI interaction. kL 'Williams, E.G., Knox, R.B. (eds) Pollination '84 Plant Cell Biology Research Centre, University of Melbourne. pp 88-97. Anderson, M.A., Cornish, E.C., Man, S.E., Williams, E.G., Hoggart, R-, Atkinson, 4., Bonig, I., Grego, 8., Simpson, R., Roche, P.J., Haley, J.D., Penshow, J.D., Niall, H.D., Tregear, G.W., Coghan, J.P., Crawford, R.J. and Clarke, A.E. (1986). Cloning of cDNA for a stylar glycoprotein associated with expression of self incompaúbility in Níco tiana alata. N arur e. 321, 38-M. Ið'/ References ciæd Anderson, W.R. and Fairbanks, D.J. (1990). Molecula¡ ma¡kers: important tools for plant genetic resource cha¡acterisation. Diversíry 6, 51-53. Apostol,.B.L., Black, IV W.C., Miller,8.R., Reiter, P. and Beaty, B.J. (1993). Estimation of the number of full sibling families at an oviposition site using RAPD-PCR markers: applications to the mosquito Aedes aegypti. Theoretical and Applied Genetícs 86, 991-1000. Arnold, M.L., Buckner, C.M. and Robinson, J.J. (1991). Pollen-mediated introgression and hybrid speciation in Lousiana irises. Proceedings of the National Academy of Sciences usÁ. 88, 1398-1402. Ayre, D.J. & Whelan, R.J. (1989) Factors controlling fruit set in hermaphroditic plants: studies with the Australian proteaceae.Tree 4,267-n2. Ayre, D.J. and V/helan, R.J. (1989). Factors controlling fruit set in hermaphroditic plants: studies with the Australian Proteaceae . Trends ín Ecology and Evolution 41267 - 272. Bachman, K. (1994). Molecular markers in plant ecology lreview]. New Phytologist. 126,403-41.8. Baverstock, P.R. and Moritz, C. (1991). Sampling design. In Molecular Systematics. Hillis, D.M., Mortiz, C. (Eds.), Sunderland, MA: Sinauer Associates,pp.13-24. Belbin, L. (1991). PATN Reference Manual. CSIRO Division of Wildlife and Ecology, Canberra. 188 References cited Bellani, L.M. and Bell, P.R. (1936). Cytoplasmic differences between the pollen grains of two cultivars of Malus domestica Borkh. correlated with viabitity and germination. Annals of Botany.58, 563-568. Bousquet, J.S., Strauss, S.H., Doerksen, A.H. and Price, R.A. (1992). Extensive va¡iation in evolutionary rate of rbcL gene sequences among seed plants. Proceedings of the National Academy of Science USA 89, 7844-7848. Brewbaker, J.L. (1959). Biology of the angiosperm pollen grain. Indian Journal of Genetics and Plant Breeding 19,l2l-133. Brits, G.J. (1985). Scarlet Ribbon - The first pastel coloured pincushion culúvar. Horticultural Science 3, 31. Brits, G.J., Jacobs, G. and Vogts, M.M. (1983). Domestication of fynbos Proteaceae as a floricultural crop. B othalia 14, 641-&6. Brooker, M.LH. (1979). A revision of the informal series Foecundae Pryor and Johnson of the genus Eucalyptus L'Herit. and notes on variation in the genus. Brunonia 2, I25- 170. Brown, A.H.D. (1979). Enzyme potymorphisms in plant populations. Theorerical P opulation B iolo gy 15, I-42. Brown, A.H.D. (1939). Genetic characterisation of plant mating systems. In: Brown, A.H.D., Clegg, M.T., Kahler, A.L. and'Weir, B.S. (eds) Plant populatíon genetics, breeding and genetic resources, Sinauer Associates, Sunderland, pp. 145-162- Ið9 References cited Brown, R. (1810). 'Prodromus Florae Novae Hollandiae et Insularae Van-Diemen.' (Taylor: London) Brown, R. (1330). 'Supplementum Primum Prodromi Florae Novae Hollandiae.' (Taylor: London.) Brummer, E.C., Bouton, J.H. and Kochert, G. (1995). Analysis of annual Medicago species using RAPD markers. Genome 38,362-367. Byrne, M., Moran, G.F. and Murrell, J.C. (1994). Detection and inheritance of RFLPs in Eucalyptus nitew. Theoretical and Applíed Genetics 89,397 -402. Cai, Q., Guy, C. and Moore, G.A. (1992). Genetic mapping of of random amplified polymorphic DNA (RAPD) markers in Citrus. Plant Genome I Abstracts: 20, San Diego, CA. Campos, L.P., Raelson, J.V. and Grant, W.F. (1994). Genome relationships among Lotus species based on random amplified polymorphic DNA (RAPD). Theoretical and Applied Genetics 88, 4I7 -422. Carlson, J.E., Tulsieram, L.K., Glaubitz, J.C., Luk, V.W.K., Kauffeldt, C. and Rutledge, R. (1991). Segregation of random amplified DNA markers in Fl progeny of conifers. Theoretical and Applied Genetics 83, 194-200. Ca¡olin, R. (1961). Pollination of Proteaceae. Australian Journal of Natural History 13, 37r-374. Carpenter, F.L. (1978) Hooks for mammal pollination. Oecologia35,L23-L32. r9u References cited Carr, 8.L., Gregory, D.P., Raven, P.H. and Tai, W. (1983). Experimental hybridisation, chromosomal diversity and phylogeny within Gaura (Onagraceae). Amertcan Journnl of Botany 75, 484-495. Carthew, S.M. (1993). Population genetic structure of Banl 566-573. Carthew, S.M., Ayre, D.J. and 'Whelan, R.J. (1988). High levels of outcrossing in populations of Banlcsia spinulosaR.Br. andBanl of Botany 36,217-223. Casper, B.B. (1988). Evidence for selective embryo abortion in Cryptantha flava. American N aturalist. 132, 3t8-326. 'W. Chalmers, K.J., Sprent, J.I., Simons, 4.J., V/augh, R. and Powell, (1992)' Patterns of genetic diversity in a tropical tree legume (Gliricídia) revealed by RAPD markers. H eredíry, 69, 465-472. Chapa:ro, J.X., Durham, R.E., Moore, G.A. and Sherman, W.B. (1987)- Use of ízozyme techniques to identify peach x "Non Pa¡eil" almond hybrids. HortScience 22, 300-302. Chase, M.W., Soltis, D.E., Olmstead, R.G., Morgan, D., Les, D.H-, Mishler, 8.D., Duvall, M.R., Price, R.4., Hillis, H.G., Qiu, Y.L., Klon, K.4., Rettig, J.H., Conti, E., Palmer, J.D., Manhart, J.R., Sytsma, K.J., Michaels, H.J., Kress, W.J., Karol, K.G., Clark, W.D., Hedren, M., Gaut, B.S., Jansen, R.K., Kim, K.J., Wimpee, C.F., Smith, J.F., Furnier, G.R., Strauss, S.H., Xiang, Q.Y., Plunkett, G.M., Soltis, 'Williams, P.S., Swensen, S.M., S.E., Gadek, P.A., Quinn, C.J., Eguiarte, L.E., Golenberg, E.M., Learn, G.H., Graham, Jr., S.'W., Barrett, S.C.H., Dayanandan, S. 191 References ciæd and Albert, V.A. (1993). Phylogenetics of seed plants: an analysis of nucleotide sequences from the plastid gene rbcL. Annals of the Míssourí Botanical Garden 80, 528- 580. Chong, D.K.X., Yang, R.C. and Yeh, F.C. (L994). Nucleotide divergence between populations of trembling aspen (Poplus temuloides) estimated with RAPDs. Current Genetics.26, 374-376. Clegg, M.T. (1980). Measuring plant mating systems. Bíoscience 30, 814-818. Clegg, M.T. (1989). Molecular diversity in plant populations.In: Plant Population Genetics, Breeding and Genetic Resources (eds Brown, H.D., Clegg, M.T., Kahler, A.L. and 'Wier, B.S.), pp. 98-115. Sinauer, Sunderland. Clegg, M.T. and Durbin, M.L. (1990). Molecular approaches to the study of plant biosystematics. Australían Systematic Botany 3, 1-8. Clegg, M.T. and Zurawski, G. (1991). Chloroplast DNA and the study of plant phylogeny. In: Soltis, P.S., Soltis, D.E., Doyle, J.J. (eds) Molecular systematics of plants. New York, London: Chapman andHall, 1-13. Clegg, M.T., Learn, G.H. and Goldenberg, E.M. (1991). Molecular evolution of chloroplast DNA. In: Selander, R.K., Cla¡k, 4.G., Whittam, T.S. (eds) Evolution at the molecular /evel. Sunderland: Sinauer, I35 -I49. Clifford, H.T. (1993). The relative fertilities of Banlcsia robur, B. oblongíþlia and their putative hybrid in southeast Queensland. Proceedíngs of the Royal Sociery of QueenslandlÙS, 13-16. r92 References cited Clifford, S.C. and Sedgley, M. (1993). Pistit structure of Banksia menziesii R. Br (Proteaceae) in relation to fertility. AwtralianJourrtnl of Botany 41,48t-490. Coates, D.J. and Sokolowski, R.E.S. (1992). The mating system and patterns of genetic variation tn Banlcsia cuneata A.S. George (Proteaceae). Heredity 69,II-20. Collins, B.C. and Spice, J. (1936) Honeyeaters and pollination biology of B. prionotes ( P roteac e ae ) Australían J ournal of B otany. 34, 17 5 -I85 - Collins, B.G. and Rebelo, T. (1987). Pollination biology of the Proteaceae in Australia and southern Africa. AustralianJournal of Ecology 12,387-42I- Couch, J.A. and P.J. Fritz. (1990). Isolation of DNA from plants high in polyphenolics. Plant Molecular Biology Reporter 8,8-12. Cowling, R.M. and Lamont, B.B. (1986) Seed release: the role of wet dry cycles. Australian f ournal of Ecology 10,169-171. Dawson, I.K., Chalmers, K.J., Waugh, R. and Powell, W. (1993). Detection and analysis of genetic variation i¡ Hordeum spontaneum populations from Israel using RAPD markers. Molecular Ecology 2,157-159. Demeke, T., Adams, R.P. and Chibba¡, R. (1992). Potential taxonomic use of random amplifred polymorphic DNA (RAPD) - a case study in Brassica.Theoretical and Applied Genetics 84,990-994. DeNettancourt, D., Deverux, M., Bozzini, 4., Cresti, M., Pacini, E. and Sarfatti, G. (1973). Ultrastructural aspects of self incompatability mechanisms in Lycopersícon peruvíanum. Journal of CeII Science 12,403-412. t93 References cited Dettmann, M.E. (1973). Angiosperrnous pollen from Albain to Turonian sediments of eastern Australia. Specíal Publicatior* Geology Sociery of Australia4,3-34. Dickenso.n, H.G. and Lewis , D. (L973). The formation of tryphine coating the pollen grains of Raphanur, and its properties relating to the self incompatibility system. Proceedings ofthe Royal Sociery ofLondon, Series B 184,148-165. Dixon, K.'W., McCredie, T.A. and Sivasithamparom, K. (1985) Grafting Banksias to avoid root rot. Awtralian H orticulture 83, 7 5 -7 9 - Do, N. and R.P. Adams. (1991). A simple technique for removing plant polysaccharide contaminants from DNA. B iotechnique s 10, 1 63- 1 66. Doust, A. (1933). 'The evolution and biogeography of Banlcsia L.f.' BSc (Hons) Thesis, The University of Sydney. Doyle, J.J. and J.L. Doyle. (1988). Isolation of plant DNA from fresh tissue. Focus 12, 13-15. Drake, D.W. (1980). Contrasting success of natural hybridisation in two Eucalyptus species pairs. Austalian Journal of Botany 28, 167 -I92. Dumas, C. and Knox, R.B. (1983). Callose and determination of pollen viability and incomp atib tlity . T he or eti c aI and Ap pli e d G e netí c s 67, I - I0 - Dunemann, F., Kahnau, R. and Schmidt, H. (1994). Genetic relationships in Malus evaluared by RAPD fingerprinting of cultiva¡s and wild species. Plant Breeding-Zeítschrift fur Pflanzenzuchtung. 113, 150- 159. 194 References cited Eltis, M.F., Sedgley, M. and Gardner, J.A. (1991). Interspecific pollen-pistil interaction ln Eucalyprus L'Her. (Myrtaceae): the effect of taxonomic distance. Annals of Botany 68, 185-194. Erickson, L.R., Straus, W.D. and Beversdorf, W.D. (1983). Resriction patterns reveal origins of chloroplast genomes in Br¿ssica amphiploids. Theoretical and Applied Genetícs 65,201-206. Ernst, A. (1953). 'Basic numbers' und Potyploidie und ihre Bedeutung Fur das Heterostylie - Problem. Arch. Julius, Klaus - Stifiung28,l-I59. Excoffier, L., Smouse, P.E. and Quattro, J.M. (L992). Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction sites. G enetic s L3l, 47 9 -49 1. Fang, G., S. Hammar and R. Grumet. (1992). A quick inexpensive method for removing olysaccharides from plant genomic DNA. Biotechniques L3,52-55. Felsenstein, J. (1981). Evolutionary trees from DNA sequences: a maximum likelyhood approach. J ournal of M ole cular Evolution 17, 368-37 6. Felsenstein, J. (1985). Confidence limits on phylogenies: an approach using the bootstrap. Evolution. 39, 7 83-7 91. Ferris, C., Oliver, R.P., Davy, 4.J., and Hewitt, G.M. (1993). Native oak chloroplasts reveal and ancient divide accross Europe. Molecular Ecology 2,337-3M. r95 References ciæd Franken, J., Custers, J.B.M. and Bino, R.J. (1938). Effects of temperature on pollen tube growth and fruit set in reciprocal crosses between Cucumís satívus and C. metuliftrus. Plant Breeding 100, 150-153. Frirz, N.K. and Hanneman, R.E. Jr. (1989). Interspecifrc incompatibility due to stylar barriers in tuber bearing and closely related non-tuber-bearing Solanums. Sexual Plant Repro ductío n 2, I84-I92. Frye, M.S. and Hedges, S.B. (1995). Monophyly of the order Rodentia inferred from mitochondrial DNA sequences of the genes for 12s IRNA, 16s rRNA, and tRNA-valine. M olecular Biology and Evolution. 12, 168-17 6. Fuss, A.M. and Sedgley, M. (1990) Floral initiation and development in relation to time of flowering in Banksia coccinea R. Br. and,Banksía menziesii R. Br. (Proteaceae) Australian Journal of Botany 38, 487-500. Fuss, A.M. and Sedgley, M. (1991) The development of hybridization techniques for Banksia menziesíi for cut flower production. Journal of Horticulrural Science 66,357- 36s. Fuss, A.M. and Sedgley, M. (1991). Pollen tube growth and seed set of Banksia coccinea R. Br. (Proteaceae). Annals of Botany 68,377-384- Fuss, A.M. and Sedgley, M. (1991). Variabitity in cut flower production of. Banksia coccinea R.Br. a¡d Banksia menziesii R. Br. at six locations in southern Australia. Australían Journal of Experimental Agrículrure 31, 853-858- Fuss, A.M. and Sedgley, M. (1991). Banksias, improving their cutflower potential. Australian H orticulrur e 89, 46-49. 196 References cifed Fuss, 4.M., Pattison, S.J., Aspinall, D. and Sedgley, M. (1992). Shoot glowth in reiation to cut flower production of Banl S cientia H orticulturae 49, 323-334. Gardner, C.A. (1928). Description of Banksia audax and Dryandra erythrocephala Journal of Proceedings of the Royal Society WA.46,I29-I33- Gardner, C.A. (1964). Contributiones Florae Australiae Occidentalis. Iournal of Proceedings of the Royal Society W.A.47, 54-59. George, A. S. (1981). The genus Banlcsia L.f. (Proteaceae) Nuytsia3,239-473. George, A. S. (19SS). New taxa and notes on Banksia L.f. (Proteaceae) Nuytsía 6, 309-317. George, A.S. (1987) The Banksia book. Kangaroo Press. The Society for Growing Australian Plants- NSW Ltd. Australia. Gielly, L. and Taberlet, P. (1994). The use of chloroplast DNA to resolve plant phylogenies: Non codong versus rbcl sequences. Molecular Biology and EvolutionLl, 769-777. Gielly, L., Yuan, Y.M., Kupfer, P. and Taberlet, P. (1996). Phylogenetic use of non coding regions in the genus Gentianal.: Chloroplast trnL (UAA) innon versus nuclear ribosomal internal transcribed spacer sequences. Molecular Phylogenetícs and Evolution 5,460-466. r97 References cited Goldenberg, E.M., Clegg, M.T., Durbin, M.L., Doebley, J. and Ma, D.P. (1993)' Evolution of a non coding region of the chloroplast genome. Molecular Phylogenetics and Evolution2,52-64. Goldingay, R.L., Schibeki, S.M. and Walker, B.A. (1991). Breeding system and pollination levels of. Banlcsia ericiþlia. Austalian Journal of Botany 39,365-372- Gower, J. C. (1935). Measures of simila¡ity, dissimilarity and distance. In Klotz, S. and Johnson, N.L. (eds). Encyclopedia of Statístical Sciences, Wiley, New York. Vol 5. pp 397-405. Grant, V. (1949). Pollination systems as isolating mechanisms in angiosperms. Evofurton 3, 82-97. 'Werner, Grattapaglia, D., Chaparro, J., Wilcox, P., McCord, S-, D., Amerson, H., McKeand, S., Bridgewater, F., Whettetr, R., O'Malley, D. and Sederoff, R. (1992). Mapping in woody plants with RAPD markers: application to breeding in forestry and horticulture . Applications of RAPD technology to Plant Breeding. Joint Plant Breeding Symposia Series, Minneapolis pp. 37-40. Gray, J.E., McCIure, B.A., Bonig, I., Anderson, M.A. and Clarke, A.E. (1991). Action of the sryle product of the self incompatability gene of Nicotiana ø/arø (sRNase) on in vitro-grown pollen tubes. Plant Ce\\.3,271-283. Griffin, A.R. and Hand, F.C. (1979). Post -anthesis development of flowers of Eucalyptus regnans F.Muell. and the timing of a¡tifrcial pollination. Australian Forestry Research 9,9-15. r9E References ciæd Griffin, 4.R., Burgess, I.P. and Wolf, L. (1988). Natural hybridisation in the genus Eucalyprus- a review. Australían Journal of Botany 36,41-66. Hamrick, J.L and Godt, M.J. (1989). Atlozyme diversity in plant species. In: Brown, A.H.D., Clegg, M.T., Kahler, A.L. and Weir, B.S. (eds) Plant population genetics, 43 - br e e ding and g e neti c r e S our c e S, Sinauer As sociates, S underland, pp. 63. Hamrick, J.L. (1983). The distribution of genetic variation within and among natural plant populations. In: Genetics and Conservation (eds C.M. Schonewald-Cox, S-M. Chambers, B. MacBride and Thomas, L.) pp. 343-348. Benjamin-Cummings, London. Hanley, B.A. and Schuler, M.A. (1988). Plant intron sequences: evidence for distinct groups of introns. Nucleic Acids Research L6,7159-7 L7 5 - Harrison, B.J. and Darby, L. (1955). Unilateral hybridisation. Nature (London).176' 982. Hawkins, J.D. (1988). A survey on intron and exon length. Nucleíc Acíds Research 16, 9893-9908. Henero, M. and Dickinson, H.G. (1981). Pollen tube development in Petuniø hybrida following compatible and incompatible intraspecific matings. Journal of CelI Science 47, 36s-383. Heslop - Harrison, J., Knox, R.B. and Heslop - Harrison, Y. (1974). Pollen wall proteins; exine held fractions associated with the incompatibility response in Cruciferae. Theoretical and Applied Genetics 44,I33-t37. r99 References cited Hestop-Harrison, J. and Heslop-Harrison, Y. (1970). Evaluation of pollen viability by enzymatically induced fluorescence: intracellular hydrolosis of fluorescein diacetate. Stain T echnolo gies 45, Ll5-I20. Heslop-Harrison, J., Heslop-Harrison, Y. and Shivanna, K.R. (1984). The evaluation of pollen quality, and a further appraisal of the fluorochromatic (FCR) test procedure. Theoretical and Applied Genetics 67,367-75. Hillis, D.M., Larson, 4., Davis, S.K. and Zimmet, E.A. (1990). Nucleic acids II: sequencing In: D.M. Hillis, C. Moritz (Eds.) Molecular Systematics. Sinauer Associates Inc., Massachusetts pp. 318-370. Hogenboom, N.G. (1973). A model for incongruity in intimate partner relationships. E up hy ric a. 22, 2I9 -233. Hogenboom, N.G. (1975). Incompatability and incongruity, two different mechanisms for the non functioning of intimate partner relationships. Proceedings of the Royal Academy of Science London, Series B 188, 36I-375. Hogenboom, N.G. (1984). Incongruity: non-functioning of inter cellular and intra cellular partner relationships through mismatching information. In Encyclopedia of Plant Physiology, New Series vol.17, Cellular Interactions, eds, H.F. Linskens and J. Heslop-Harrisonm, pp 640-65 1. Springer-Verlag, Berlin. Hoot, S.8., Reznicek, A.A. and Palmer, J.D. (1994). Phylogenetic relationships in Anemone (Ranunculaceae) based on morphology and chloroplast DNA. Systematic Botarry 19, t69-200. 200 References cited Howell, E.C., Newbury, H.J., Swennen, R.L., Withers, L.A. and Ford-Llyod' B.V. (lgg4). The use of RAPD for identifying and classifying Musa germplasm . Genome 37,328-332. Hsiao, C., Chatterton, N.J., Asay, K.H. and Jensen, K.B. (1995). Molecular phylogeny of the Pooídeae (Poaceae) based on nuclear rDNA (ITS) sequences. Theoretical and Applied Generics 90, 389-398. Huff, D.R., Peakall, R. and Smouse, P.E. (1993). RAPD variation within and among natural populations of outcrossing buffalograss ÍBuchloe dacryloides (Nutt.) Engelm.l Theoretícal ant Applied Genetícs 86,921-934. Jaccard, P. (1901). Etude comparative de la distribution florale dans une portion des Alpes et des Jura. Bull. Soc. Vau.doíse Sci. Nat.37, 547-579- Jackson, D.4., Somers, K.M. and Harvey, H.H. (1989). Similarity coefficients: Measures of co-occurence, and association or simply measures of occurence? American N aturalis t I33, 436-453. Jahnen, W., Batterham, M.P., Clarke, 4.E., Moritz, R.L. and Simpson, R.J. (1989). Identification, isolation and N-terminal sequencing of style glycoproteins associated with self-incompatability in N icotiann al anta. P lant C eII. l, 493-499. John, M.E. (1992). An efficient method for isolation of RNA and DNA from plants containing polyphenolics. N ucleic Actds Res earch 20, 238I - Johnson, L.A.S. and Briggs, B.G. (1975). On the proteaceae - The evolution and classification of a southern family. Botanical Journal of the Linnaean Socíery 83-182. 20r References cited Johnson, L.A.S. and Briggs, B.G. (1963). Evolution in the Proteaceae. Australían Journal of Botany IL,2I-6I. Johnson,.L.A.S. and Briggs, B.G. (1981). Three old southern families - Myrtaceae, proteaceae and Restionaceae. ln'Ecological biogeography of Australía'. (Ed. A- Keast') pp.427-469. (W. Junk: The Hague). Johnson, L.A.S. and Briggs, B.G. (1933). Proteaceae. In 'Flowering Plants in Australia'. (Eds. B.D. Morley and H.R. Toelken.) pp.238-244. (Rigby: Adelaide)' Kamboj, R.K. and Jackson, J.F. (1986). Self-incompatibility alleles control a low molecular weight, basic protein in pistils of Petunia hybrida. Theoretícal and Applied Genetics 71, 815-819. Kapoor, M.L. and Sharma, V.K. (1984). Hybrids between Eucalyptus citriodora Hook. and E. torellianaF.v.Muell. in India. S tvlae G e netica - 33' 42-46. Knox, R.B., Willing, R.R., Pryor, LD. (L972)- Interspecific hybridisation of poplars using recognition pollen. Silvae Genetica.2L, 65-69. Kuo, J., Hocking, P.J. and Pate, J.S. (1982). Nutrient reserves in seeds of selected proteaceous species from south western Australia. Australian Journal of Botany 30,231- 249 Ladiges, P.Y., Volovicic, F. and Drinnan, A.N. (1995). Eucalypt Phylogeny - Molecules and morphology. Awtralian Systematic Botany 8,483-497 . 202 References ciæd Lamonr, B. (1935). Fire responses of sclerophyl shrublands - a population ecology approach with particular reference to the genus Banksialn Fíre ecology and managemeru of WA ecosystems (J. R. Ford ed Bendley, WA Institute of Technology pp 4I-48). Lamont, B.B. and Ba:rett, G.J. (198S). Constraints on seed production and storage in a root-suckenng Banlcsia. Journal of Ecology 76, 1069-1082- Langton, F.A. (1987).Breeding for improved ornamental plants In Improving Vegetatively Propagated Crops. Academic hess Limited. London. Lewis, J. and Bell, D.T. (19S1). Reproductive isolation of co-occvring Banksiø species at the Yule Brook Botany Reserve, Western Australia. AustalianJournal of Botany 29, 665-74. Linskens, H.F. (1965). Biochemistry of incompatability. Proc. 1lth Intern. Congr. Genet. The Hague,1963In'Genetics today', Geerts, S.J. (Ed.), 3,621-636- Liu, Z and Furnier, G.R. (1993). Comparison of allozyme, RFLP and RAPD markers for revealing genetic variation within and between trembling asPen and bigtooth aspen. Theoretical and Applied Genetics 87, 97-105. Lodhi, M.A., Reisch, B.I. and'Weeden, N.F. (1992). Molecular genetic mapping and genome size. Plant Genome Abstracts;37, San Diego, CA- Lynch, M. and Milligan, B.G. (1993). Analysis of polulation genetic structuro with RAPD markers. Molecular Ecolo gy. 3, 91-99. Maddison, D.R. (1991). The discovery and importance of multiple islands of most parsimonious trees. Systematic Zoololgy 40, 315-328. 203 References cited Maguire, T.L., Sedgley, M. and Conran, J.G. (1996). Banksia sec|.. Coccinea (A.S' George) T. Maguire et aI., (Proteaceae). A new section. Australian Systematic Botany (n press). Maguire, T. L. and Sedgley, M. (1997). Interspeciflc and intergeneric pistil - pollen compatibility with B anlcs i a c o c cine a R.Br. (Proteaceae). (in pres s)' Maguire, T.L., Collins, G.G. and Sedgley, M. (1994). A modified CTAB DNA exrraction procedure for plants belonging to the family Proteaceae. Plant Molecular Biology Reporrer 12, 106-109. Mailer, R.J., Scarth, R. and Fristensky, B. (1994). Discrimination rimong cultivars of rapeseed (Brassíca napus L.)using DNA polymorphisms amplified from arbitrary primers. Theoretical and Applied Genetics 87,697-704. Maniatis, T., Fritsch, E.F. and Sambrook, J. (1938) Molecular cloníng A lab manual Cold Spring Harbour. New York. Martin, H.A. (1978). Evolution of the Australian Flora and vegetation through the Tertiary: Evidence from pollen . Alcheringa. 2, I8l-202. Mattson, O., Knox, R.B., Heslop-Harrison, J. and Heslop-Harrison, Y- (1974) P¡otein pellicle of stigmatic papillae as a probable regulatory site in incompatability reactions. Nature L47,298-300. Mau, S.L., Raff, J. and CIúke, A.E. (1982). Isolation and partial characterisation of components of Prunus avium L. styles, including an antigenic glycoprotein associated with a self-incompatability genotype. P lanta. 156, 505-5 1 6. 204 Refe¡ences ciæd McBride, L.J., Koepf, S.M., Gibbs, R.4., Salser, W., Mayrand, P'E', Hunkapiller, M.V/. and Kronick, M.N. (1989). Automated DNA sequencing methods involving polymerase chain reaction. C Ii ni c al C hemi s try 3 5, 219 6-220 t - McConchie, C.A., Vithange, V. and Batten, D.J. (1994). Intergeneric hybridisation between Litchi (Litchí chinensís Sonn.) and longan (Dimocarpus longan Lour.). Annals of Botany 74, ll1-118. McCourt, R.M. (1995). Green algal phylogeny [review]. Trends in Ecology and Evolutíon. 10, 159- 163. Meissner, C.F. (1856). Proteaceae. In "Prodromus Systematis Naturales Regní Vegetabilis, Vol. 14.' (Ed. A.P. de Candolle.) pp.45l-467. (Victor Masson: Paris.) Meskimen, G. and Franklin, E.C. (1984). Hybridity in the Eucalyptus grandis breeding population in Florida. Resea¡ch Paper No. SE - 242. South eastern Forest Experiment Station, USDA Forest Service. Moran, G.F. and Hopper, S.D. (1987). Conservation of the genetic resources of rare and. widespread eucalypts in remnant vegetation.In: Nature Conservation: The Role of Remnants of NativeVegetation (eds D.A. Saunders, G.W. Arnold, A.A. Burbridge and A.J.M. Hopkins) pp. 151-162. Surrey Bearry and Sons, Chipping Norton. Morgan, D.R. and Soltis, D.E. (1993). Phylogenetic relationships among members of Saxifragacaeae sensu lato based on rbcL sequence data. Annals of the Missouri Botanical Gardens 80,631-660. 205 References cited Mummenhoff, K. and Koch, M. (1994). Chloroplast DNA restriction site variation and phylogenetic relationships in the genus Thlaspi sensu lato (Brassicaceae). Systematic Botany 19, 73-88. Nasrallah, J.B., Kao, T.H., Goldberg, M.L. and Nasrallah, M.E. (1985). A cDNA clone encoding an S-locus-specific glycoprotein from Brassica oleracea. Nature 318, 263-267. Nei, M. and Li, W.H. (1979). Mathematical model for studying genetic variation in terms of restriction endonucleases. Proceedings of the National Academy of Science u.s.A.74,5267-5273. Nelson, M. and M. McClelland. (1991). Site specific methylation: effect on DNA modification methyltransferases and restriction endonucleases. Nucleic Acids Research 19,2045-2070. Nesbitt, K.4., Potts, B.M., Vaillancourt, R.E., West, A.K. and Reid, J.B. (1995). Partitioning and distribution of RAPD va¡iation in a forest tree species, Eucalyptus globulus (Myrtaceae). Heredity.74, 628-637 . Nicoletti, V.G and Condorelli, D.F. (1993). Optimised PEG method for rapid plasmid DNA purification: High yield from 'midi - prep'. BioTechniques 14,532-536. Norrhcote, K.H. (1979). Afactorial keyfor the recognitíon of Australíansoils, Adelaide, SA, 124 pp. Olmstead, R.G. and Sweere, J.A. (1994). Combining data in phylogenetic systematics - an empirical approach using three molecular sets in the Solanaceae. Systematic Biology 43, 467-48t. zu6 References ciæd Palmer, J.D. (1937). Chloroplast DNA evolution and biosystematic uses of chloroplast DNA variation. Amerícan N arur alist 130, S 6-529. Palmer, J.D., Jansen, R.K., Michaels, H.J., Chase, M.W. and Manhaft, J.R. (1988). Chloroplast DNA variation and plant phylogeny. Annals of the Missouri Botanical Garden. 75, 1180-1206. Parfitt, D.E., Arulsekar, D.E. and Ramming, D.W. (i985). Identification of Plum x Peach hybrids by isozyme analysis. H orts cie nce. 20, 246-248 - Parvin, P. (19S1). Hybrídisation of proteas in Hawaü. First International Conference of Protea Growers: 65-67. Payne, C.D. (1987). Genstat 5 reference manual. Oxford University Press, Oxford. Peakall, R., Smouse, P.E. and Huff, D.R. (1995). Evolutionary implications of allozyme and RAPD variation in diptoid populations of dioecious buffalograss Buchloe dacryloides. Molecular Ecology 4, 135-L47 . phillips, R.B., Manley, S.A. and Daniels, T.J. (1994). Systematics of the Salmonoid genus Salvelinus infered from ribosomat DNA sequences. Canadian Journal of Fisheries and Aquatic Scíences.51, 198-204. Potts, B.M. and Jackson, W.D. (1989). Evolutionary processes in the Tasmanian high altitude eucalypts. In: Flora and Fauna of Alpine Austalia: Ages and Origins. (Ed.) B. Ba¡low. pp. 511-527. CSIRO, Melbourne. Potrs, B.M. and Reid, J.B. (1985). Analysis of a hybrid swarm between Eucalyptus risdonäHook.f. andE. amygdalinalabilI. AustralianJournalof Botany39,497-504- 207 References cited Ports, B.M. and Reid, J.C. (1988). Hybridisation as a dispersal mechanism. Evolutíon. 42, t245-r255. Potrs, B.M. and Reid, J.R. (1qS3). Hybridisation between Eucalyptus obliqua L'Herit. and E. pulchella. AustralianJournal of Botany 3l,2ll-229. Potts, B.M., Potts, V/.C. and Cauvin, B. (1987). Inbreeding and interspecific hybridis ati on in Eucalyptus gunnii. S ilv ae G enetica 36, 19 4-L99. Prakash, N. (1986). 'Methods in Plant Mícrotechnique'2nd. Ed. (Botany Department, University of New England, Armidale, NSV/). Pryor, L.D. and Willing, R.R. (1974). The production of interspecific hybrids between incompatible pairs of eucalypts.Interrational Plant Propagators Socíery 24,265-268. Raff, J.W. (1983). Flower bud development on stone fruit varieties. Aøsrralian H ortic ultur al R e s e ar ch N ew sI ett er. 55, 35-3 6. Raff, R.4., Marshall, C.R. and Turbeville, J.M. (1994). Using DNA sequences to unravel the Cambrian radiation of the animal phyla lreview]. Annual Review of Ecology and Systematics. 25, 351-37 5. Ramsey, H.P. (1963). Chromosome numbers in the Proteaceae- Australian Journal of Botany LI, I-20. Ramsey, M. and Vaughton, G. (1991) Self incompatibility, protandry, pollen production and pollen longevity in Banlcsia menziesü. Australían Journal of Botany 39, 497 -504. 208 References cited Ramsey, M.W. (1936). Aspects of the anthecology of. Banksia menzíesii R.Br. and Banksia littoralis R.Br. (Proteaceae) in relation to fruit set. B.Sc. Hons. Thesis, Murdoch University, Perth. Rieseberg, L.H., Beckstrom-Sternberg, S. and Doan, K. (1990). Helíanthw annus ssp. texanus has chloroplast DNA and nuclear ribosomal genes of. Helíanthus debiliJ ssp. cucumeríþlius. Proceedings of the National Academy of Sciences USA.87, 593-597 . Ritland, K. (1983). Estimation of mating systems. In: Isozymes in Plant Genetícs and Breeding. (eds S.D. Tanksley and T.J. Orton.) pp. 289-302. Elsevier Science Publ., Amsterdam. Ritland, K. and Clegg, M.T. (1987). Evotutionary analysis of plant DNA sequences. Americ an N aturalist 130, S74-S 100. Rodgers, D.J. and Tanimoto, T.T. (1960). A computer program for classifying plants. Science 132, 11 15-1 1 18. Rohl, L.J., Fuss, 4.M., Dhaliwal, J.4., Webb, M.G. and Lamont, B.B. (1994). Investigation of flowering in Banksia baxterí R.Br. and B. hookeríana Meissner for improving pruning practises. Australían Journal of Experimental Agriculture.34, 1209- 1216. Rossetto, M., Weaver, P.K. and Dixon, K.W. (1995). Use of RAPD analysis in devising conservation strategies for the rare and endangered Grevillea scapigera (hoteaceae). M olecular Ecolo gy. 4, 321-329. Rouse, J.L., Knox, R.B. and V/illiams, E.G. (1985). Unilateral hybridisation in Rhododendron. In S exual Reproduction ín S e ed P lants, F erns and M oss es . P roc eeding s of 2rJ9 References cited the 8th International Symposíum on Sexual Reproductíon in Seed Plants, Ferns and Mosses. lgS4,Wageninen. ed.M.T.M. Willemse and J.L. van Went. pp. 115-16. Pudoc, Wageningen. Ruanro, G. and Kidd, K.K. (1991). Coupled amptification and sequencing of genomic DNA. Proceedings of the Natíonal Academy of Sciences USA. 88, 2815-2819. Ruggeri, C. (1959). Alcune osservazioni sulla va¡iabilita de caratteri morfologiei di Eucalyptus x tabutti Vilmorin in F1 ed F2. Pubblicazioni del Centro dí Sperimentazione Agricole e Forestale Roma 2, 203-22I. Russell, J.R., Hosein, F., Johnson, E., Waugh, R. and Powell, W. (1993). Geneúc differentiation of cocoa (Theobroma caca.o L.) poputations revealed by RAPD analysis. Molecular Ecology 2, 89-97 . Russell, P.F. and Rao, T.R. (1940). On habitat and association of species of anopheline larvae in south-eastern Madras. Journal of the Malar. Institute of India.3, 153-178. Sale, M.M., Potts, B.M., West, A.K. and Reid, J.B. (1996). Relationships within Eucalyptus (Myrtaceae) using PCR - amplifrcation and southern hybridisation of c hloroplas t DNA. Aus tr alian Sy s t e matí c B o t any 9, Z7 3 -282. Salkin, A. (1937) Breeding systems in eastern Australian banksias. Victorian Naturalist 103, 43-47. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989). Molecular Cloning: A Laboratory Manual,2 nd edn. Cold Spring Harbor Laboratory Press, New York. z\) References ciled Sampson, J.F., Cotlins, B.G. and Coates, D.J. (1994). Mixed mating ln Banksia browníiBaxter ex R.Br. (Proteaceae). AustralianJourral of Botany 42,I03-L1'l- Sang, T.,.Crawford, D.J., Stuessy, T.F. O MS (1995). ITS sequences and the phylogeny of the genus Robinsonia (Asteraceae). Systematic Botany 20, 55-&. Schaal, B.A., O'Kane-Jr, S.L. and Rogstad, S.H. (1991). DNA variation in plant population s. Trends in Ecology and Evolutíon 6,329-333. Schemske, D.W. and Lande, R. (1935). The evolution of self fertilisation and inbreeding depression in plants. II Empirical observations. Evolution 39,41-52. Schnell, R.J., Ronning, C.M. and Knight, R.J. (1995). Identifrcation of cultivars and validation of genetic relationship s in Mangifera indica L. using RAPD ma¡kers. Theoretical and Applied Genetícs 90,269-274. Schoen, D.J. (1982). The breeding system of GiIía achilleiþIi¿: variation in floral cha¡acteristic s and outcros sing rate. Evolutio n. 36, 59 6-6L3. Scotr, J.K. (1980). Estimation of the outcrossing rate for Banksia attenuata R.Br. and Banksia menzíesii R.Br. @oteaceae). Australian Journal of Botany 28,53-59. Sears, B.B. (1980). Elimination of plastids during spennatogenesis and fertilisation in the plant kingdom. P lasmíd. 4, 233-25 5. Seavey, S.R. and Bawa, K.S. (1986). Late acting self incompatibility in Angiosperms. The Botanical Review.52, 195-219. Sedgley, M. (1981). Storage of avocado pollen. Euphytica.30, 595-599. 2tL References cifed Sedgley, M. (1933) Pollen tube growth ln Macadamia. Scientia. Horticulturae.lS, 333- 34r. Sedgley,.M. and Griffin, A.R. (1989). Sexual reproduction of tree crops. Academic Press, London. Sedgley, M. and Ha¡bard, J. (1993). Pollen storage and breeding system in relation to controlled pollination of four species of Acacia (Iæguminosae: Mimosoideae). Australian Journal of Botany 41, 601-609. Sedgley, M. and Smith, R.M. (1989). Pistil receptivity and pollen tube growth in relation to the breeding system of Eucalyptus woodwardiMaiden (Symphyomyrtus). Botanical Gazzete (in press). Sedgley, M., Harbard., J., Smith, R.M. and W'ickneswari, R. (1992). Development of hybridisation techniques for Acacia mangium and Acacia aurículíþrmís.In'Breeding Technologíes for Tropical Acacias'. Australian Centre for International Agricultural Research Proceedings No. 37. (Eds L.T. Carron and K. Aken.) pp. 63-9. (ACIAR:Canberra.). Sedgley, M., Sierp, M., Wallwork, M.4., Fuss, A.M. and Thiele, K. (1993). Pollen presenter and pollen morphology of Banksia L.f. (Proteaceae) Australian Journal of Botany 41, 439-64. Sedgley, M., Sierp, M.G. and Maguire, T.L. (1994). Interspecific hybridisation involving Banksia prionotes Lind. and Banksía menziesíi R.Br. (Proteaceae). Internatíonal Journal of Plant Sciences t55,755-762. 2t2 References ciæd Sedgley, M., Wirthensohn, M. and Fuss, A.M. (1991) Selection and breeding of 'Western Banksias. 1n International Protea Association, 6th Biennial Conference. Perth Australia. 193-272. Sedgley, M., Wirthensohn, M.G. and Delaporte, K. (1996). Interspecific hybridisation betlveen Banksía hookeríana Meisn and B. prionotes Lindl. (Proteaceae). International Journnl of Plant Sciences (in press). Shivanna, K.R. and Heslop-Harrison, J. (1981). Membrane state and pollen viability. Annals of Botany 47,759-770. Smith - White, S. (1959). Cytological evolution in the Australian flora. Cold Springs Harbour Symposium on Quantatative Biology 24,273-289 Smith, L.M., Sanders, J.2., Kaiser, R.J., Hughes, P., Dodd, C., Connell, C'R', Heiner, C., Kent, S.B.H. and Hood, L.E. (1986). Fluorescence detection in automated DNA sequence analysis. Nature 321, 674-679. Stammers, M., HarriS, J., Evans, G.M., Hayward, M.D. and FOrSter, J.W. (1995). Use of random PCR (RAPD) technoiogy to analyse phylogenetic relationships in the Loliuml F estuca complex. H er edity 7 4, 19 -27 . Stanley, R. and Kirby, E.G. (1973). In Kozlowski (Ed) Shedding of plant parts. p-29. Academic Press, New York. Stanley, R.G. and Linskens, H.F. (I974). PoIIen: biology, bíochemistry and managemenr. Springer, Berlin Heidelberg New York. 2r3 References ciæd Stephenson, A.G. and Bertin, R.I. (1983). Male competition female choice and sexual selection in plants. In: Pollínatíon Biology. (ed) L. Real. pp 109-49. Academic Press, Orlando. Stock, W.D., Pate, J.S., Kuo, J. and Hansen, A.P. (1989). Resource control of seed set in Banksia laricína C. Ga¡dner (Proteaceae). Functional Ecology 3,453-460. Swofford, D.L. (1990). PAUP: Phylogenetic analysis using parsimony. Computer progam distributed by the Illinois Natural History Survey, Champaign,Illinois. Swofford, D.L. and Olsen, G.J. (1990). Phylogeny reconstruction. In: Hillis, D.M. and Moritz, C. (Eds.) Molecular Systematics. Sinuaer Associates, Sunderland, Massachusetts pp.411-501. Sybenga, J. (1983) Genetic manipulation in plant breeding: somatic versus generative Thoretícal and Applied Generics 66,179-20t. Taberlet, P., Gielly, L., Pautou, G. and Bouvet, J. (1991). Universal primers for amplification of three non-coding regions of chloroplast DNA. Plant Molecular Biology. 17, 1105-1109. i" frop.r "ro\ a Taji, A. and Williams, R (1990) *repoÉati€È of Australian plants by tissue culture. Australían Horticulture 88, 89-92. Taylor, A. and Hopper, S. (1988). The Banl No. 8. Australian Government Publishing Service, Canberra. Thiele, K. (1993). A cladistic analysis of Banksia L.f. (PhD Thesis, School of Botany: University of Melbourne). 214 References cited Vasil, I. and Vasil, V. (1930). Isolation and culture of protoplasts. In: Perspectives in Plant Cell andTissue Culture. Ed. I. Vasil. International Review of Cytology 118. pp. 1-19. Academic Press, New York. Venkata Rao, C. (1971) Proteaceae 207 pp CSIR, New Dehli, India. Virk, P.S., Ford-Lloyd, B.V., Jackson, M.T. and Newbury, H.J. (1995). Use of RAPD for the study of diversity within plant germplasm collections. Heredíry 74,I70- r79. Voss, H., Schwager, C., Wirkner, U., Sporat,8., Zimmermann, J., Rosenthal, 4., Erfle, H., Stegemann, J. and Ansorge, W. (1989). Direct genomic fluorescent on-line sequencing and analysis using ín vitro amplification of DNA. Nucleíc Acíds Research L7 , 2517-2517. 'Walker, B.A. and Whelan, R.J. (1991). Can andromonoecy explain low fruit:flower ratios in the Proteaceae? Biological Jourral of the Linnnen Sociery 44,4I-46. 'Wallace, D.D. and O'Dowd, D.J. (1989) The effects of nutrients and inflorescence damage by inscets on fruit settn Banksia spínulosa. Oecologia79,482-488. 'Walrop, A.B. (1983) The opening mechanism of folicles of some species of Banksia. Australian Journal of Botany 31, 485-500. Waugh, R. and Powell, W. (1992). Using RAPD ma¡kers for crop improvement.Trends in Biotechnology 10, 186-191. zL5 References cited Weining, S. and Langridge, P. (1991). Identification and mapping of polymorphisms in cereals based on the polymerase chain reaction. Theoretícal and Applíed Genetics 82, 209-2t6. Welsh, J. and McClelland, M. (1990). Fingerprinting genomes using PCR with arbitrary primers. N uc e i c A cids R e s e ar ch. 18, 7 213 -7 218. Vy'elsh, J., Honeycutt, R.J., McClelland, M. and Sobral, B.V/.S' (1991). Parentage dtermination in maize hybrids using the arbitrarily primed polymerase chain rection (AP- PCR). Theoretical and Applied Genetics 82,473-476. 'Whelan, R.J. (1994). Conserving nature's ways. Today's Life Science.6,20-25. Whitehouse, H.L.K. (1950). Muttiple - allelo - morph incompatibility of pollen and style in the evolution of the angiospeÍns. Annals of Botany New Series 14, 198-21-6. Wier, B. (1991). Intraspecific differentiation. In Molecular Systematics. Hillis, D.M., Mortiz, C. (Eds.), Sunderland, MA: Sinauer Associates,pp.3T3-405. V/ilikie, S.E., Issac, P.G. and Slater, R.J. (1993). Random amplif,red polymorphic DNA (RAPD) markers for genetic analysis in AIIium. Theoretical and Applied Genetics 87, 668-672. V/illiams, E.G., Knox, R.B. and Rouse, J.L. (1982). Overgrowth of pollen tubes in embryo sacs of R/¿adodendronfollowing interspecific pollinations. AustralianJournal of Botany 34, 4t3-423. 'Williams, E.G., Rouse, J.L., Palser, B.F. and Knox, R.B. (1990). Reproductive biology of Rhododendron. Horticultural Reviews L2, t-67 . 216 References ciæd Williams, J.G.K., Kubelik, 4.R., Livak, K.J., Rafalski, J.A. and Tingey, S.V. (1990). DNA polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucleic Acids Research 18, 653 1-6535. Williams, J.G.K., Kubelik, 4.R., Livak, K.J., Rafalski, J.A. and Tingley, S.V. (1990). Genetic analysis using random amplified polymorphic DNA markers. Methods in E nzy mo I o gy 2I8, 7 04-7 40. Witkowski, E.T.F., Lamont, B.B. and Connel, S.J. (1991). Seed bank dynamics of three co-occuring banksias in south coastal western Australia: the role of plant age, cockatoos, senescence and interfi¡e establishmeît. Australian Journal of Botany 39, 385- 397. Wodehouse, R.P. (1935). Pollen grains. New York. McGraw Hill. Wolfe, K.H., Sharp, P.M. and Li, W-H. (1989). Rates of synonomous substitution in plant nuclear genes. Journal of Molecular Evolution29,208-2II. Xiang, Q.Y., Soltis, D.E., Morgan, D.R and Soltis, P.S. (1993). Phylogenetic relationships of Cornas L. sensu lato and putative relatives inferred from rbcL sequence datz. Annals of the Missouri Botanícal Gardens 80,723-734. Yates, I.E. and Sparks, D. (1989). Hydration and temperature influence in vítro germination of pecan pollen. Journal of the American Sociery of Horticultural Science 114, 599-605. Yu, K. and Paulis, K.P. (1992) Optimisation of the PCR program for RAPD analysis. Nucleíc Acíds Research 20,2606. '¿L',l References cited Yu, K. and Pauls, K.P. (1993). Rapid estimation of genetic relatedness among heterogenous populations of alfalfa by random amplification of bulked genomic DNA samples. Theoretical and. Applíed Genetícs 86,788-794. Zammit, C. and Westoby, M. (1983) Pre-dispersal and losses, and the survival of seeds and seedlings of two serotinoas Banksi¿ shrubs in burnt and unbumt heath. Journal of Ecology 76,200-214. 2I8 Aonendix Appendix Maguire, T.L., Collins, G.G. and Sedgley, M. (1994). A modified CTAB DNA extraction proceedure for plants belonging to the family koteaceae. Plant Molecular Biology Reporter 12, 106-109. 2t9 PLANT MOLECULAR BIOLOGY REPORTER Volume 12, Numb er 2, June 1994 Maguire, T. L., Collins, G. G. & Sedgley, M. (1994). A modified CTAB DNA extraction procedure for plants belonging to the family proteaceae. Plant Molecular Biology Reporter, 12(2), 106-109. NOTE: This publication is included in the print copy of the thesis held in the University of Adelaide Library. It is also available online to authorised users at: http://dx.doi.org/10.1007/BF02668371 Pollination t4 Consequences of Different Reproductive Modes 15 Self incompatibility 16 Stigma T7 Stylar inhibition 18 Ovary inhibition 18 Evolution of self incompatibility 18 The self incompatibility mechanism 19 Interspecific Incomp atibilitY L9 Reproductive EcologY 20 Reproductive isolation 22 Natural Hybúdisation 22 Hybrid fitness 24 Pattems of hybridisation 25 Chromosome Studies 25 Propagation 25 Commercial Importance and Improvement of B attksia 26 Interspecifrc Hybridis ation 29 Table of Contents